https://chiosa.app/papers/habit-bench-evaluating-habit-induction-from-longitudinal-weak-evidence-in-agent-memory-2610-24520Chiosaen2026-10-28T19:27:59ZMemory systems struggle to turn user history into sound choiceshttps://chiosa.app/papers/hy-nomas-towards-free-form-interactive-world-generation-2610-36729Chiosaen2026-10-28T19:27:59ZA video model carries scenes and actions across turnshttps://chiosa.app/papers/memory-tax-dissecting-the-net-value-of-agent-memory-2610-34073Chiosaen2026-10-28T19:27:59ZMany agent memory systems lose to direct history searchhttps://chiosa.app/papers/representation-redistribution-in-sparse-autoencoders-2610-21240Chiosaen2026-10-28T19:27:59ZA method finds feature changes that recur across promptshttps://chiosa.app/papers/spatial-audio-generation-and-editing-with-executable-plans-and-self-distilling-transfusion-2610-34415Chiosaen2026-10-28T19:27:59ZOne framework plans and edits spatial audio sceneshttps://chiosa.app/papers/work-agents-systems-capabilities-and-optimization-2610-35845Chiosaen2026-10-28T19:27:59ZA framework connects digital work agents to the work they must finishhttps://chiosa.app/papers/worldcast-world-consistent-character-video-generation-with-large-angle-camera-control-2610-33135Chiosaen2026-10-28T19:27:59ZWorldCast keeps characters coherent through large camera moveshttps://chiosa.app/papers/balanced-state-subspaces-for-mamba-2-state-compression-2610-36617Chiosaen2026-10-28T15:18:30ZMamba-2 compression preserves jointly useful state directionshttps://chiosa.app/papers/benchmeup-building-software-engineering-benchmarks-from-natural-language-specifications-2610-37283Chiosaen2026-10-28T15:18:30ZBenchMeUp turns research specifications into runnable benchmarkshttps://chiosa.app/papers/can-discrete-tokens-serve-as-semantically-lossless-representations-of-visual-signals-2610-37662Chiosaen2026-10-28T15:18:30ZA visual tokenizer keeps discrete features informativehttps://chiosa.app/papers/can-discrete-tokens-serve-as-semantically-lossless-representations-of-visual-signals-2610-37662?lang=itChiosait2026-10-28T15:18:30ZUn tokenizer visivo mantiene informative le caratteristiche discretehttps://chiosa.app/papers/can-test-time-memory-evolution-outperform-test-time-training-for-scientific-discovery-2610-32046Chiosaen2026-10-28T15:18:30ZExternal memory helps models find better mathematical solutionshttps://chiosa.app/papers/channel-selection-for-mixture-of-experts-decoding-needs-no-calibration-data-2610-37386Chiosaen2026-10-28T15:18:30ZFrozen MoE decoding can select channels without calibrationhttps://chiosa.app/papers/channel-selection-for-mixture-of-experts-decoding-needs-no-calibration-data-2610-37386?lang=itChiosait2026-10-28T15:18:30ZLa decodifica MoE congelata può selezionare canali senza calibrazionehttps://chiosa.app/papers/channel-selection-for-mixture-of-experts-decoding-needs-no-calibration-data-2610-37386?lang=frChiosafr2026-10-28T15:18:30ZLe décodage MoE figé peut sélectionner les canaux sans calibrationhttps://chiosa.app/papers/chronoreplay-one-answer-interface-is-not-enough-for-streaming-spatial-intelligence-2610-34283Chiosaen2026-10-28T15:18:30ZChronoReplay trains video models to answer spatial questionshttps://chiosa.app/papers/closed-loop-aware-planning-via-imitation-free-speed-scoring-for-end-to-end-autonomous-driving-2610-36079Chiosaen2026-10-28T15:18:30ZA driving model learns to follow requested speedshttps://chiosa.app/papers/codecap-bench-can-code-serve-as-a-good-video-caption-2610-37341Chiosaen2026-10-28T15:18:30ZExecutable code can preserve details in video captionshttps://chiosa.app/papers/codecap-bench-can-code-serve-as-a-good-video-caption-2610-37341?lang=itChiosait2026-10-28T15:18:30ZIl codice eseguibile può conservare dettagli nelle descrizioni videohttps://chiosa.app/papers/codecap-bench-can-code-serve-as-a-good-video-caption-2610-37341?lang=esChiosaes2026-10-28T15:18:30ZEl código ejecutable puede conservar detalles en los pies de videohttps://chiosa.app/papers/codecap-bench-can-code-serve-as-a-good-video-caption-2610-37341?lang=frChiosafr2026-10-28T15:18:30ZLe code exécutable peut préserver les détails des légendes vidéohttps://chiosa.app/papers/concordat-typed-sessions-for-llm-agents-2610-37461Chiosaen2026-10-28T15:18:30ZA protocol gate stopped unconfirmed airline changes under pressurehttps://chiosa.app/papers/crisp-co-designing-pruning-and-scheduling-for-efficient-moe-inference-on-edge-servers-2610-37442Chiosaen2026-10-28T15:18:30ZCRISP coordinates pruning and scheduling for edge MoEhttps://chiosa.app/papers/do-we-need-an-llm-to-know-why-sources-conflict-decision-models-for-knowledge-conflict-typing-in-rag-2610-36915Chiosaen2026-10-28T15:18:30ZMost tested models match an older language model at classifying search conflictshttps://chiosa.app/papers/efficient-quantization-training-with-stochastic-rounding-via-structured-subspace-randomness-2610-36587Chiosaen2026-10-28T15:18:30ZShared rounding thresholds speed training when errors are correctedhttps://chiosa.app/papers/evolving-agent-teams-2610-36848Chiosaen2026-10-28T15:18:30ZA supervisor helps coding teams rethink GPU kernelshttps://chiosa.app/papers/featurearc-benchmarking-coding-agents-on-evolving-features-2610-37434Chiosaen2026-10-28T15:18:30ZCoding agents often lose feature correctness across stageshttps://chiosa.app/papers/follow-the-gradient-tracing-safety-degradation-back-to-training-data-2610-37481Chiosaen2026-10-28T15:18:30ZGradient filtering removes risky training exampleshttps://chiosa.app/papers/follow-the-gradient-tracing-safety-degradation-back-to-training-data-2610-37481?lang=itChiosait2026-10-28T15:18:30ZIl filtraggio basato sui gradienti rimuove esempi rischiosi dai dati di addestramentohttps://chiosa.app/papers/follow-the-gradient-tracing-safety-degradation-back-to-training-data-2610-37481?lang=esChiosaes2026-10-28T15:18:30ZEl filtrado por gradientes retira ejemplos de entrenamiento riesgososhttps://chiosa.app/papers/gas-pfn-graph-aware-spatiotemporal-prior-data-fitted-networks-for-zero-shot-traffic-forecasting-2610-37598Chiosaen2026-10-28T15:18:30ZA graph-aware model forecasts traffic without retraininghttps://chiosa.app/papers/gas-pfn-graph-aware-spatiotemporal-prior-data-fitted-networks-for-zero-shot-traffic-forecasting-2610-37598?lang=esChiosaes2026-10-28T15:18:30ZUn modelo con grafos predice el tráfico sin reentrenarsehttps://chiosa.app/papers/gas-pfn-graph-aware-spatiotemporal-prior-data-fitted-networks-for-zero-shot-traffic-forecasting-2610-37598?lang=frChiosafr2026-10-28T15:18:30ZUn modèle tenant compte des routes prévoit le trafic sans réentraînementhttps://chiosa.app/papers/governed-sovereign-rag-for-public-administration-tripartite-epistemic-routing-facet-guided-clarification-and-2610-36561Chiosaen2026-10-28T15:18:30ZA government assistant checks generated numbers against procedureshttps://chiosa.app/papers/heads-of-security-learned-attention-gates-improve-prompt-injection-robustness-2610-37657Chiosaen2026-10-28T15:18:30ZLearned attention gates reduce prompt injection in two open modelshttps://chiosa.app/papers/heads-of-security-learned-attention-gates-improve-prompt-injection-robustness-2610-37657?lang=itChiosait2026-10-28T15:18:30ZGate di attenzione appresi riducono la prompt injection in due modelli open-weighthttps://chiosa.app/papers/heads-of-security-learned-attention-gates-improve-prompt-injection-robustness-2610-37657?lang=esChiosaes2026-10-28T15:18:30ZCompuertas de atención aprendidas reducen la inyección en dos modelos abiertoshttps://chiosa.app/papers/heads-of-security-learned-attention-gates-improve-prompt-injection-robustness-2610-37657?lang=frChiosafr2026-10-28T15:18:30ZDes portes d’attention apprises réduisent l’injection d’instructions dans deux modèles ouvertshttps://chiosa.app/papers/how-much-can-best-of-n-help-a-world-action-model-measuring-selectable-headroom-and-realizing-it-in-closed-loo-2610-37196Chiosaen2026-10-28T15:18:30ZRobot action selection gains depend on candidate designhttps://chiosa.app/papers/hqclethe-streaming-hqc-encapsulation-for-tiny-stateless-servers-2610-36808Chiosaen2026-10-28T15:18:30ZHqcLethe streams HQC keys without per-client memoryhttps://chiosa.app/papers/infradev-can-coding-agents-develop-ai-infrastructure-and-how-can-we-tell-2610-36589Chiosaen2026-10-28T15:18:30ZHidden checks expose coding agents’ incomplete infrastructure changeshttps://chiosa.app/papers/know-then-draw-a-new-learning-paradigm-for-visual-generation-2610-37626Chiosaen2026-10-28T15:18:30ZA two-stage method separates visual learning from renderinghttps://chiosa.app/papers/know-then-draw-a-new-learning-paradigm-for-visual-generation-2610-37626?lang=esChiosaes2026-10-28T15:18:30ZUn método de dos etapas separa el aprendizaje visual de la generaciónhttps://chiosa.app/papers/know-then-draw-a-new-learning-paradigm-for-visual-generation-2610-37626?lang=frChiosafr2026-10-28T15:18:30ZUne méthode en deux étapes sépare apprentissage visuel et renduhttps://chiosa.app/papers/learning-visuomotor-quadrotor-flight-through-dynamic-narrow-gaps-via-differentiable-simulation-2610-37557Chiosaen2026-10-28T15:18:30ZQuadrotors learn to cross moving gaps in simulationhttps://chiosa.app/papers/learning-visuomotor-quadrotor-flight-through-dynamic-narrow-gaps-via-differentiable-simulation-2610-37557?lang=itChiosait2026-10-28T15:18:30ZI quadrotori imparano in simulazione ad attraversare varchi mobilihttps://chiosa.app/papers/linmix-data-mixing-as-linear-domain-attribution-2610-37390Chiosaen2026-10-28T15:18:30ZVertex sampling helps mix many language-model data domainshttps://chiosa.app/papers/lossless-compression-of-proximity-graph-indexes-for-efficient-approximate-nearest-neighbors-search-2610-37176Chiosaen2026-10-28T15:18:30ZHNSW graph compression cuts storage with little latency costhttps://chiosa.app/papers/mle-world-predicting-machine-learning-experiment-outcomes-with-a-learned-world-model-2610-37175Chiosaen2026-10-28T15:18:30ZPredicted program feedback helps agents search for solutionshttps://chiosa.app/papers/multi-agent-reinforcement-learning-of-co-adaptive-hip-exoskeleton-control-marl-exo-for-metabolic-energy-reduc-2610-37441Chiosaen2026-10-28T15:18:30ZA hip exoskeleton lowers walking energy versus a passive device in a small testhttps://chiosa.app/papers/multi-state-linear-attention-with-generalized-second-order-routing-2610-37645Chiosaen2026-10-28T15:18:30ZMerlin helps long-context models weight useful segmentshttps://chiosa.app/papers/multi-state-linear-attention-with-generalized-second-order-routing-2610-37645?lang=itChiosait2026-10-28T15:18:30ZMerlin aiuta i modelli a contesto lungo a pesare i segmenti utilihttps://chiosa.app/papers/multi-state-linear-attention-with-generalized-second-order-routing-2610-37645?lang=esChiosaes2026-10-28T15:18:30ZMerlin ayuda a modelos de contexto largo a ponderar segmentos útileshttps://chiosa.app/papers/multi-state-linear-attention-with-generalized-second-order-routing-2610-37645?lang=frChiosafr2026-10-28T15:18:30ZMerlin aide les modèles à long contexte à pondérer les segments utileshttps://chiosa.app/papers/pailitaommsearch2-0-technical-report-scaling-generative-retrieval-to-unified-end-to-end-multimodal-search-2610-36958Chiosaen2026-10-28T15:18:30ZOne model handles three kinds of shopping searchhttps://chiosa.app/papers/pdrvla-a-vision-language-action-model-with-physical-dynamics-representations-2610-37381Chiosaen2026-10-28T15:18:30ZA robot model learns scene changes without predicting videohttps://chiosa.app/papers/pdrvla-a-vision-language-action-model-with-physical-dynamics-representations-2610-37381?lang=itChiosait2026-10-28T15:18:30ZUn modello robotico apprende i cambiamenti della scena senza generare videohttps://chiosa.app/papers/pdrvla-a-vision-language-action-model-with-physical-dynamics-representations-2610-37381?lang=esChiosaes2026-10-28T15:18:30ZUn modelo robótico aprende cambios de escena sin predecir videohttps://chiosa.app/papers/pdrvla-a-vision-language-action-model-with-physical-dynamics-representations-2610-37381?lang=frChiosafr2026-10-28T15:18:30ZUn modèle robotique apprend l’évolution des scènes sans générer de vidéohttps://chiosa.app/papers/pipelink-contract-guided-linking-of-gpu-kernel-components-2610-37200Chiosaen2026-10-28T15:18:30ZPipeLink lets separately built GPU kernel parts work togetherhttps://chiosa.app/papers/pneuma-a-large-scale-real-world-time-series-benchmark-for-commercial-aircraft-predictive-maintenance-2610-37563Chiosaen2026-10-28T15:18:30ZA public aircraft dataset exposes limits in early fault detectionhttps://chiosa.app/papers/pneuma-a-large-scale-real-world-time-series-benchmark-for-commercial-aircraft-predictive-maintenance-2610-37563?lang=itChiosait2026-10-28T15:18:30ZUn dataset pubblico sugli aerei mostra i limiti del rilevamento precoce dei guastihttps://chiosa.app/papers/pneuma-a-large-scale-real-world-time-series-benchmark-for-commercial-aircraft-predictive-maintenance-2610-37563?lang=esChiosaes2026-10-28T15:18:30ZUn conjunto público de datos de aeronaves muestra límites en la detección temprana de falloshttps://chiosa.app/papers/pneuma-a-large-scale-real-world-time-series-benchmark-for-commercial-aircraft-predictive-maintenance-2610-37563?lang=frChiosafr2026-10-28T15:18:30ZUn jeu de données public révèle les limites de la détection précoce des pannes d’avionhttps://chiosa.app/papers/pushing-dexterity-and-safety-for-in-hand-object-pose-reaching-that-works-remarkably-well-2610-37652Chiosaen2026-10-28T15:18:30ZA robot hand reaches poses while dropping fewer objectshttps://chiosa.app/papers/pushing-dexterity-and-safety-for-in-hand-object-pose-reaching-that-works-remarkably-well-2610-37652?lang=itChiosait2026-10-28T15:18:30ZUna mano robotica raggiunge le pose facendo cadere meno oggettihttps://chiosa.app/papers/pushing-dexterity-and-safety-for-in-hand-object-pose-reaching-that-works-remarkably-well-2610-37652?lang=esChiosaes2026-10-28T15:18:30ZUna mano robótica alcanza posiciones y deja caer menos objetoshttps://chiosa.app/papers/pushing-dexterity-and-safety-for-in-hand-object-pose-reaching-that-works-remarkably-well-2610-37652?lang=frChiosafr2026-10-28T15:18:30ZUne main robotique atteint des poses en faisant moins tomber les objetshttps://chiosa.app/papers/rendering-free-event-frame-simulation-for-fast-reinforcement-learning-of-autonomous-robot-navigation-2610-37543Chiosaen2026-10-28T15:18:30ZA rendering-free simulator trains drone navigation policies in minuteshttps://chiosa.app/papers/rendering-free-event-frame-simulation-for-fast-reinforcement-learning-of-autonomous-robot-navigation-2610-37543?lang=itChiosait2026-10-28T15:18:30ZUn simulatore senza rendering addestra in pochi minuti i controllori di navigazione dei dronihttps://chiosa.app/papers/rendering-free-event-frame-simulation-for-fast-reinforcement-learning-of-autonomous-robot-navigation-2610-37543?lang=esChiosaes2026-10-28T15:18:30ZUn simulador sin renderizado entrena la navegación de drones en minutoshttps://chiosa.app/papers/rendering-free-event-frame-simulation-for-fast-reinforcement-learning-of-autonomous-robot-navigation-2610-37543?lang=frChiosafr2026-10-28T15:18:30ZUn simulateur sans rendu entraîne la navigation des drones en quelques minuteshttps://chiosa.app/papers/resple2-inertial-informed-recursive-non-uniform-spline-estimation-for-multi-sensor-odometry-with-global-refin-2610-37602Chiosaen2026-10-28T15:18:30ZRESPLE2 predicts robot motion from inertial datahttps://chiosa.app/papers/resple2-inertial-informed-recursive-non-uniform-spline-estimation-for-multi-sensor-odometry-with-global-refin-2610-37602?lang=itChiosait2026-10-28T15:18:30ZRESPLE2 predice il movimento dei robot con dati inerzialihttps://chiosa.app/papers/resple2-inertial-informed-recursive-non-uniform-spline-estimation-for-multi-sensor-odometry-with-global-refin-2610-37602?lang=esChiosaes2026-10-28T15:18:30ZRESPLE2 predice el movimiento del robot con datos inercialeshttps://chiosa.app/papers/roboauto-autonomous-robot-program-search-through-multi-level-physical-understanding-2610-36533Chiosaen2026-10-28T15:18:30ZRoboAuto grounds robot programs in physical scene knowledgehttps://chiosa.app/papers/rollback-experience-backtracking-in-world-models-for-policy-improvement-2610-36068Chiosaen2026-10-28T15:18:30ZRollback helps robots learn by revisiting key actionshttps://chiosa.app/papers/scaling-laws-for-tool-use-agents-via-controllable-environment-synthesis-2610-37648Chiosaen2026-10-28T15:18:30ZMore varied tool training data helps as budgets growhttps://chiosa.app/papers/secdmp-provably-secure-data-memory-dependent-prefetching-extended-version-2610-37454Chiosaen2026-10-28T15:18:30ZA hardware defense keeps prefetchers from scanning secret datahttps://chiosa.app/papers/select-ttt-query-conditioned-selective-fast-weight-updates-for-in-place-test-time-training-2610-36794Chiosaen2026-10-28T15:18:30ZSelect-TTT directs memory updates toward relevant texthttps://chiosa.app/papers/spine-compiling-subtask-trees-into-structured-context-for-long-horizon-coding-agents-2610-37224Chiosaen2026-10-28T15:18:30ZSpine helps coding agents track nested workhttps://chiosa.app/papers/spine-compiling-subtask-trees-into-structured-context-for-long-horizon-coding-agents-2610-37224?lang=itChiosait2026-10-28T15:18:30ZSpine aiuta gli agenti di programmazione a seguire il lavoro annidatohttps://chiosa.app/papers/teach-reality-not-ethics-a-drop-in-continued-pretraining-corpus-installs-coupled-human-biosphere-ai-welfare-d-2610-37484Chiosaen2026-10-28T15:18:30ZTraining helps models resist pressure to approve ecological harmhttps://chiosa.app/papers/tessera-structure-completed-evidence-units-for-web-retrieval-augmented-generation-2610-37080Chiosaen2026-10-28T15:18:30ZTessera gets webpage evidence into the reader’s contexthttps://chiosa.app/papers/the-structure-gap-when-more-data-cannot-replace-structured-reasoning-2610-37523Chiosaen2026-10-28T15:18:30ZKnown task structure can resolve gaps left by exampleshttps://chiosa.app/papers/the-structure-gap-when-more-data-cannot-replace-structured-reasoning-2610-37523?lang=itChiosait2026-10-28T15:18:30ZLa struttura nota del compito può colmare i vuoti lasciati dagli esempihttps://chiosa.app/papers/the-trinity-capability-driven-recursive-self-improvement-for-multimodal-agents-2610-36077Chiosaen2026-10-28T15:18:30ZAgents can improve their tools by grouping shared abilitieshttps://chiosa.app/papers/thinkworld-world-imagination-reinforces-chain-of-dynamics-reasoning-2610-37666Chiosaen2026-10-28T15:18:30ZRobot learning uses video to ground motion reasoninghttps://chiosa.app/papers/thinkworld-world-imagination-reinforces-chain-of-dynamics-reasoning-2610-37666?lang=itChiosait2026-10-28T15:18:30ZI video aiutano i robot a ragionare sul movimentohttps://chiosa.app/papers/thinkworld-world-imagination-reinforces-chain-of-dynamics-reasoning-2610-37666?lang=frChiosafr2026-10-28T15:18:30ZLes vidéos aident les robots à raisonner sur le mouvementhttps://chiosa.app/papers/tops-reshaping-flow-source-and-transport-for-multi-task-robot-manipulation-2610-36694Chiosaen2026-10-28T15:18:30ZTOPS improves robot task success with task-specific flow pathshttps://chiosa.app/papers/trident-seq-profile-matched-prefix-future-contrast-for-label-efficient-payment-risk-modeling-2610-37365Chiosaen2026-10-28T15:18:30ZMatching payment histories helps risk models learn from fewer labelshttps://chiosa.app/papers/trident-seq-profile-matched-prefix-future-contrast-for-label-efficient-payment-risk-modeling-2610-37365?lang=itChiosait2026-10-28T15:18:30ZAbbinare le cronologie di pagamento aiuta i modelli di rischio con meno etichettehttps://chiosa.app/papers/trident-seq-profile-matched-prefix-future-contrast-for-label-efficient-payment-risk-modeling-2610-37365?lang=esChiosaes2026-10-28T15:18:30ZEmparejar historiales de pago ayuda a aprender con menos etiquetashttps://chiosa.app/papers/vibevideobench-evaluating-multimodal-agents-on-code-driven-video-generation-2610-35864Chiosaen2026-10-28T15:18:30ZA new test shows where video-making agents still fall shorthttps://chiosa.app/papers/vision-language-action-or-world-action-just-action-models-2610-37219Chiosaen2026-10-28T15:18:30ZRobot policies can learn from the future without generating images (F1, F3)https://chiosa.app/papers/watermark-basins-black-box-detection-of-distilled-llms-2610-37266Chiosaen2026-10-28T15:18:30ZA text-only test traces watermarked training into student modelshttps://chiosa.app/papers/when-does-perturbation-consistency-add-information-beyond-softmax-confidence-theory-and-evidence-for-the-deci-2610-36640Chiosaen2026-10-28T15:18:30ZPerturbation consistency adds little to clean image error checkshttps://chiosa.app/papers/whether-is-not-when-separating-occurrence-from-timing-in-polymarket-prices-2610-36639Chiosaen2026-10-28T15:18:30ZPrediction prices reveal more about whether than whenhttps://chiosa.app/papers/workfactory-turning-professional-intents-into-executable-work-for-scalable-agent-training-2610-37161Chiosaen2026-10-28T15:18:30ZA task-building framework checks whether work tasks can be solvedhttps://chiosa.app/papers/world-dreamer-v5-lite-a-multi-robot-world-action-model-with-one-dreamer-2610-37518Chiosaen2026-10-28T15:18:30ZA shared spatial map helps robots coordinate across viewshttps://chiosa.app/papers/world-dreamer-v5-lite-a-multi-robot-world-action-model-with-one-dreamer-2610-37518?lang=itChiosait2026-10-28T15:18:30ZUna mappa spaziale condivisa aiuta i robot a coordinarsihttps://chiosa.app/papers/your-policy-is-secretly-a-harness-the-decision-geometry-of-agent-interfaces-2610-37255Chiosaen2026-10-28T15:18:30ZA harness can hurt a frozen agent and help it after retraininghttps://chiosa.app/papers/at-at-automatic-tool-optimization-from-agent-traces-2610-34679Chiosaen2026-10-27T15:32:57ZAT-AT improves shared tools from agents’ past runshttps://chiosa.app/papers/auto-research-arena-benchmarking-and-advancing-auto-research-in-language-modeling-2610-35866Chiosaen2026-10-27T15:32:57ZA test compares language-model research across nine goalshttps://chiosa.app/papers/auto-research-arena-benchmarking-and-advancing-auto-research-in-language-modeling-2610-35866?lang=itChiosait2026-10-27T15:32:57ZUn test confronta la ricerca sui modelli linguistici su nove obiettivihttps://chiosa.app/papers/continual-learning-exam-benchmarking-long-horizon-rsi-and-experiential-learning-in-llm-agents-2610-33956Chiosaen2026-10-27T15:32:57ZA test of whether agents carry experience forwardhttps://chiosa.app/papers/copying-as-defense-protecting-images-from-editing-by-eliciting-model-copying-2610-35566Chiosaen2026-10-27T15:32:57ZMaking image editors copy can help protect photos from editshttps://chiosa.app/papers/copying-as-defense-protecting-images-from-editing-by-eliciting-model-copying-2610-35566?lang=itChiosait2026-10-27T15:32:57ZSpingere gli editor di immagini a copiare può proteggere le foto dalle modifichehttps://chiosa.app/papers/copying-as-defense-protecting-images-from-editing-by-eliciting-model-copying-2610-35566?lang=esChiosaes2026-10-27T15:32:57ZHacer que los editores de imágenes copien puede proteger las fotos de las edicioneshttps://chiosa.app/papers/copying-as-defense-protecting-images-from-editing-by-eliciting-model-copying-2610-35566?lang=frChiosafr2026-10-27T15:32:57ZInciter les éditeurs d’images à copier peut protéger les photos des retoucheshttps://chiosa.app/papers/correlation-aware-exploration-for-long-trace-reasoning-2610-34949Chiosaen2026-10-27T15:32:57ZInfo-GRPO links reasoning paths to training-time seedshttps://chiosa.app/papers/correlation-aware-exploration-for-long-trace-reasoning-2610-34949?lang=itChiosait2026-10-27T15:32:57ZInfo-GRPO collega percorsi di ragionamento a seed di traininghttps://chiosa.app/papers/correlation-aware-exploration-for-long-trace-reasoning-2610-34949?lang=esChiosaes2026-10-27T15:32:57ZInfo-GRPO vincula rutas de razonamiento con semillas de entrenamientohttps://chiosa.app/papers/correlation-aware-exploration-for-long-trace-reasoning-2610-34949?lang=frChiosafr2026-10-27T15:32:57ZInfo-GRPO relie les chemins de raisonnement aux seeds d’entraînementhttps://chiosa.app/papers/error-diffusion-replaces-latent-weights-training-binary-weight-language-models-on-a-few-bits-of-dithered-stat-2610-34566Chiosaen2026-10-27T15:32:57ZBinary models can train with small update buffershttps://chiosa.app/papers/hierarchical-skill-evolution-for-coding-agent-2610-35616Chiosaen2026-10-27T15:32:57ZCoding agents gain reusable repair skills while their model stays fixedhttps://chiosa.app/papers/hierarchical-skill-evolution-for-coding-agent-2610-35616?lang=itChiosait2026-10-27T15:32:57ZGli agenti di programmazione acquisiscono competenze riutilizzabili senza modificare il modellohttps://chiosa.app/papers/hierarchical-skill-evolution-for-coding-agent-2610-35616?lang=frChiosafr2026-10-27T15:32:57ZLes agents de programmation acquièrent des compétences de réparation réutilisables sans modifier leur modèlehttps://chiosa.app/papers/hifi-a-highly-efficient-framework-for-statistical-inference-with-foundation-models-2610-34550Chiosaen2026-10-27T15:32:57ZA two-stage method estimates uncertainty with foundation modelshttps://chiosa.app/papers/i-hitl-inalienable-human-in-the-loop-2610-35835Chiosaen2026-10-27T15:32:57ZHuman approval needs a trusted path beyond the agenthttps://chiosa.app/papers/i-hitl-inalienable-human-in-the-loop-2610-35835?lang=frChiosafr2026-10-27T15:32:57ZL’approbation humaine exige un canal hors du contrôle de l’agenthttps://chiosa.app/papers/janus-synthesizing-cross-layer-defenses-against-web-deserialization-attacks-2610-35532Chiosaen2026-10-27T15:32:57ZJanus coordinates web defenses against changing exploitshttps://chiosa.app/papers/janus-synthesizing-cross-layer-defenses-against-web-deserialization-attacks-2610-35532?lang=itChiosait2026-10-27T15:32:57ZJanus coordina le difese web contro exploit variabilihttps://chiosa.app/papers/janus-synthesizing-cross-layer-defenses-against-web-deserialization-attacks-2610-35532?lang=esChiosaes2026-10-27T15:32:57ZJanus coordina defensas web frente a exploits cambianteshttps://chiosa.app/papers/janus-synthesizing-cross-layer-defenses-against-web-deserialization-attacks-2610-35532?lang=frChiosafr2026-10-27T15:32:57ZJanus coordonne les défenses web contre les exploits variableshttps://chiosa.app/papers/knowing-when-the-evidence-is-enough-requirement-directed-search-with-horn-closure-certificates-2610-34684Chiosaen2026-10-27T15:32:57ZCertify checks each needed fact before answering a search questionhttps://chiosa.app/papers/prefetch-steered-routing-deterministic-cache-aware-expert-selection-for-offloaded-mixture-of-experts-decoding-2610-34721Chiosaen2026-10-27T15:32:57ZExpert routing stays stable as copy timing changeshttps://chiosa.app/papers/reva-real-time-vulnerability-detection-through-contextual-probing-of-large-language-models-2610-35829Chiosaen2026-10-27T15:32:57ZA probe reads vulnerability signals inside language modelshttps://chiosa.app/papers/reva-real-time-vulnerability-detection-through-contextual-probing-of-large-language-models-2610-35829?lang=frChiosafr2026-10-27T15:32:57ZUn probe lit les signaux de vulnérabilité au sein des modèles de langagehttps://chiosa.app/papers/sentrybench-a-cost-sensitive-benchmark-for-adversarial-robustness-of-tool-using-llm-agents-in-public-safety-d-2610-34535Chiosaen2026-10-27T15:32:57ZA trust gate cuts attacks on Haiku, with mixed results across modelshttps://chiosa.app/papers/shutdownmod-turning-any-trajectory-into-a-shutdown-evaluation-2610-35486Chiosaen2026-10-27T15:32:57ZA method tests whether agents stop when askedhttps://chiosa.app/papers/shutdownmod-turning-any-trajectory-into-a-shutdown-evaluation-2610-35486?lang=esChiosaes2026-10-27T15:32:57ZUn método evalúa si los agentes se detienen cuando se les pidehttps://chiosa.app/papers/shutdownmod-turning-any-trajectory-into-a-shutdown-evaluation-2610-35486?lang=frChiosafr2026-10-27T15:32:57ZUne méthode vérifie si les agents s’arrêtent à la demandehttps://chiosa.app/papers/the-agent-in-the-middle-how-ai-mediated-access-reshapes-web-vulnerability-exploitation-2610-35541Chiosaen2026-10-27T15:32:57ZVela shows Web interfaces shape which known attacks get throughhttps://chiosa.app/papers/the-agent-in-the-middle-how-ai-mediated-access-reshapes-web-vulnerability-exploitation-2610-35541?lang=itChiosait2026-10-27T15:32:57ZVela mostra quali attacchi noti passano attraverso le interfacce Webhttps://chiosa.app/papers/the-agent-in-the-middle-how-ai-mediated-access-reshapes-web-vulnerability-exploitation-2610-35541?lang=esChiosaes2026-10-27T15:32:57ZVela muestra cómo las interfaces web dejan pasar ataques conocidoshttps://chiosa.app/papers/the-noise-transmission-burden-of-diffusion-network-why-image-prediction-can-outperform-noise-prediction-2610-34834Chiosaen2026-10-27T15:32:57ZWhy diffusion models struggle to predict noise through bottleneckshttps://chiosa.app/papers/the-noise-transmission-burden-of-diffusion-network-why-image-prediction-can-outperform-noise-prediction-2610-34834?lang=itChiosait2026-10-27T15:32:57ZPerché i modelli di diffusione faticano a predire il rumore con input compressihttps://chiosa.app/papers/the-physical-cost-of-context-quality-constrained-resource-accounting-for-coding-agents-2610-34763Chiosaen2026-10-27T15:32:57ZHow coding-agent context changes server workhttps://chiosa.app/papers/triage-efficient-vision-language-action-inference-needs-both-reuse-and-pruning-2610-34519Chiosaen2026-10-27T15:32:57ZRobots can mix token reuse and pruning to speed up vision-language controlhttps://chiosa.app/papers/triage-skip-skim-or-refine-attention-tiles-in-video-diffusion-transformers-2610-35706Chiosaen2026-10-27T15:32:57ZTriage matches video attention work to each tilehttps://chiosa.app/papers/triage-skip-skim-or-refine-attention-tiles-in-video-diffusion-transformers-2610-35706?lang=itChiosait2026-10-27T15:32:57ZTriage adatta il calcolo dell’attenzione video a ogni tilehttps://chiosa.app/papers/triage-skip-skim-or-refine-attention-tiles-in-video-diffusion-transformers-2610-35706?lang=esChiosaes2026-10-27T15:32:57ZTriage adapta el cálculo de atención de video a cada bloquehttps://chiosa.app/papers/triage-skip-skim-or-refine-attention-tiles-in-video-diffusion-transformers-2610-35706?lang=frChiosafr2026-10-27T15:32:57ZTriage adapte le calcul de l’attention vidéo à chaque tuilehttps://chiosa.app/papers/understanding-and-repairing-cooperative-perception-under-adverse-weather-2610-35110Chiosaen2026-10-27T15:32:57ZWeather changes outside objects can disrupt cooperative detectionhttps://chiosa.app/papers/understanding-and-repairing-cooperative-perception-under-adverse-weather-2610-35110?lang=esChiosaes2026-10-27T15:32:57ZLos cambios meteorológicos fuera de los objetos pueden alterar la detección cooperativahttps://chiosa.app/papers/understanding-and-repairing-cooperative-perception-under-adverse-weather-2610-35110?lang=frChiosafr2026-10-27T15:32:57ZLes changements météo autour des objets perturbent la détection coopérativehttps://chiosa.app/papers/v-3-ast-bench-verifiable-valid-and-versatile-safety-evaluation-for-coding-agents-2610-35188Chiosaen2026-10-27T15:32:57ZA benchmark checks coding-agent safety through real effectshttps://chiosa.app/papers/v-3-ast-bench-verifiable-valid-and-versatile-safety-evaluation-for-coding-agents-2610-35188?lang=itChiosait2026-10-27T15:32:57ZUn benchmark verifica la sicurezza degli agenti di coding attraverso effetti realihttps://chiosa.app/papers/v-3-ast-bench-verifiable-valid-and-versatile-safety-evaluation-for-coding-agents-2610-35188?lang=frChiosafr2026-10-27T15:32:57ZUn benchmark évalue la sécurité des agents de code à travers leurs effets réelshttps://chiosa.app/papers/worldpager-zero-stall-tiered-memory-for-long-horizon-world-models-on-a-single-consumer-gpu-2610-35529Chiosaen2026-10-27T15:32:57ZWorldPager keeps long world-model sessions within memory limitshttps://chiosa.app/papers/worldpager-zero-stall-tiered-memory-for-long-horizon-world-models-on-a-single-consumer-gpu-2610-35529?lang=itChiosait2026-10-27T15:32:57ZWorldPager prolunga le sessioni dei world model entro i limiti di memoriahttps://chiosa.app/papers/benice-closing-looping-s-inference-memory-gap-via-kv-sharing-2610-31633Chiosaen2026-10-27T12:15:30ZTraining lets looped language models reuse inference cacheshttps://chiosa.app/papers/benice-closing-looping-s-inference-memory-gap-via-kv-sharing-2610-31633?lang=esChiosaes2026-10-27T12:15:30ZEl entrenamiento permite reutilizar cachés de inferencia en modelos de lenguaje en buclehttps://chiosa.app/papers/eat-your-precision-precision-objective-mitigates-reward-hacking-in-rubric-based-rl-2610-33512Chiosaen2026-10-27T12:15:30ZA training reward balances requested content with answer qualityhttps://chiosa.app/papers/eat-your-precision-precision-objective-mitigates-reward-hacking-in-rubric-based-rl-2610-33512?lang=itChiosait2026-10-27T12:15:30ZUna ricompensa bilancia contenuti richiesti e qualità della rispostahttps://chiosa.app/papers/eat-your-precision-precision-objective-mitigates-reward-hacking-in-rubric-based-rl-2610-33512?lang=esChiosaes2026-10-27T12:15:30ZUna recompensa de entrenamiento equilibra el contenido solicitado y la calidad de las respuestashttps://chiosa.app/papers/eat-your-precision-precision-objective-mitigates-reward-hacking-in-rubric-based-rl-2610-33512?lang=frChiosafr2026-10-27T12:15:30ZUne récompense d’entraînement équilibre contenu demandé et qualité des réponseshttps://chiosa.app/papers/how-much-does-the-harness-matter-and-for-whom-a-multi-model-study-of-context-in-llm-driven-vulnerability-disc-2610-28802Chiosaen2026-10-27T12:15:30ZCode context helps some vulnerability-finding models morehttps://chiosa.app/papers/how-much-does-the-harness-matter-and-for-whom-a-multi-model-study-of-context-in-llm-driven-vulnerability-disc-2610-28802?lang=itChiosait2026-10-27T12:15:30ZIl contesto del codice aiuta alcuni modelli a trovare vulnerabilitàhttps://chiosa.app/papers/how-much-does-the-harness-matter-and-for-whom-a-multi-model-study-of-context-in-llm-driven-vulnerability-disc-2610-28802?lang=esChiosaes2026-10-27T12:15:30ZEl contexto de código ayuda más a algunos modelos que buscan vulnerabilidadeshttps://chiosa.app/papers/how-much-does-the-harness-matter-and-for-whom-a-multi-model-study-of-context-in-llm-driven-vulnerability-disc-2610-28802?lang=frChiosafr2026-10-27T12:15:30ZLe contexte de code aide certains modèles à trouver des failleshttps://chiosa.app/papers/inside-jepa-from-prediction-to-relational-grounding-in-latent-world-models-2610-31194Chiosaen2026-10-27T12:15:30ZAction prediction gives relational channels meaning in video modelshttps://chiosa.app/papers/inside-jepa-from-prediction-to-relational-grounding-in-latent-world-models-2610-31194?lang=itChiosait2026-10-27T12:15:30ZPrevedere le azioni dà significato ai canali relazionali nei modelli videohttps://chiosa.app/papers/lean-loss-engineered-modulation-of-the-model-weight-sensitivity-distribution-for-efficient-error-tolerant-neu-2610-28872Chiosaen2026-10-27T12:15:30ZTraining can shrink the weights most vulnerable to bit flipshttps://chiosa.app/papers/lean-loss-engineered-modulation-of-the-model-weight-sensitivity-distribution-for-efficient-error-tolerant-neu-2610-28872?lang=itChiosait2026-10-27T12:15:30ZL’addestramento può ridurre i pesi più vulnerabili ai flip dei bithttps://chiosa.app/papers/lean-loss-engineered-modulation-of-the-model-weight-sensitivity-distribution-for-efficient-error-tolerant-neu-2610-28872?lang=esChiosaes2026-10-27T12:15:30ZEl entrenamiento puede reducir los pesos más vulnerables a inversiones de bitshttps://chiosa.app/papers/lifetime-valid-decisions-by-continually-evolving-ai-analysts-2610-30562Chiosaen2026-10-27T12:15:30ZA validation method controls risk as AI analysts changehttps://chiosa.app/papers/lifetime-valid-decisions-by-continually-evolving-ai-analysts-2610-30562?lang=itChiosait2026-10-27T12:15:30ZUn metodo di validazione controlla il rischio mentre cambiano gli analisti AIhttps://chiosa.app/papers/lifetime-valid-decisions-by-continually-evolving-ai-analysts-2610-30562?lang=esChiosaes2026-10-27T12:15:30ZUn método de validación controla el riesgo cuando cambian los analistas de IAhttps://chiosa.app/papers/lifetime-valid-decisions-by-continually-evolving-ai-analysts-2610-30562?lang=frChiosafr2026-10-27T12:15:30ZUne méthode de validation maîtrise le risque quand les analystes IA évoluenthttps://chiosa.app/papers/ngramrl-accelerating-rl-post-training-for-structured-generation-via-rollout-aware-n-gram-speculative-decoding-2610-32881Chiosaen2026-10-27T12:15:30ZNgramRL reuses training answers to speed reinforcement-learning rolloutshttps://chiosa.app/papers/ngramrl-accelerating-rl-post-training-for-structured-generation-via-rollout-aware-n-gram-speculative-decoding-2610-32881?lang=esChiosaes2026-10-27T12:15:30ZNgramRL reutiliza respuestas de entrenamiento para acelerar las ejecuciones de generación por refuerzohttps://chiosa.app/papers/ngramrl-accelerating-rl-post-training-for-structured-generation-via-rollout-aware-n-gram-speculative-decoding-2610-32881?lang=frChiosafr2026-10-27T12:15:30ZNgramRL réutilise les réponses d’entraînement pour accélérer les séquences générées en RLhttps://chiosa.app/papers/permit-securing-persistent-memory-for-llm-agents-via-machine-checked-integrity-theorems-and-scalable-communit-2610-28987Chiosaen2026-10-27T12:15:30ZA confirmation gate controls how agents reuse memoryhttps://chiosa.app/papers/permit-securing-persistent-memory-for-llm-agents-via-machine-checked-integrity-theorems-and-scalable-communit-2610-28987?lang=esChiosaes2026-10-27T12:15:30ZUna barrera de confirmación controla cómo los agentes reutilizan la memoriahttps://chiosa.app/papers/recertifying-decisions-via-e-closure-uniform-improvements-of-conformal-risk-control-and-randomized-e-bh-2610-35821Chiosaen2026-10-27T12:15:30ZA method turns unused risk guarantees into more informative decisionshttps://chiosa.app/papers/recertifying-decisions-via-e-closure-uniform-improvements-of-conformal-risk-control-and-randomized-e-bh-2610-35821?lang=itChiosait2026-10-27T12:15:30ZUn metodo trasforma il margine inutilizzato nelle garanzie sul rischio in decisioni più informativehttps://chiosa.app/papers/recertifying-decisions-via-e-closure-uniform-improvements-of-conformal-risk-control-and-randomized-e-bh-2610-35821?lang=esChiosaes2026-10-27T12:15:30ZUn método aprovecha garantías de riesgo sin usar para tomar decisiones más informativashttps://chiosa.app/papers/reliable-ai-assisted-decisions-via-selective-auditing-2610-31662Chiosaen2026-10-27T12:15:30ZMethods reduce expert reviews while limiting escaping AI errorshttps://chiosa.app/papers/reliable-ai-assisted-decisions-via-selective-auditing-2610-31662?lang=itChiosait2026-10-27T12:15:30ZMetodi per ridurre le verifiche degli esperti e limitare gli errori dell’IA senza controllohttps://chiosa.app/papers/retrieval-augmented-generation-compositionality-2610-29722Chiosaen2026-10-27T12:15:30ZRetrieval techniques lose gains when combinedhttps://chiosa.app/papers/sat-sampling-at-scale-via-gpu-based-boolean-constraint-propagation-2610-28805Chiosaen2026-10-27T12:15:30ZA GPU sampler finds valid circuit inputs in parallelhttps://chiosa.app/papers/separating-thinking-from-spelling-for-efficient-generative-recommendation-2610-32980Chiosaen2026-10-27T12:15:30ZA small decoder takes recommendation search off the full modelhttps://chiosa.app/papers/the-entire-quantum-computer-on-a-classical-leash-2610-31775Chiosaen2026-10-27T12:15:30ZA classical client can check quantum work and keep its outputhttps://chiosa.app/papers/accelerating-so-2-equivariant-materials-models-through-low-precision-emulation-2610-35164Chiosaen2026-10-27T09:16:52ZFlashSO(2) speeds symmetry-preserving layers with integer emulationhttps://chiosa.app/papers/accelerating-so-2-equivariant-materials-models-through-low-precision-emulation-2610-35164?lang=itChiosait2026-10-27T09:16:52ZFlashSO(2) accelera gli strati che preservano la simmetria con l’emulazione interahttps://chiosa.app/papers/adapt-for-planning-not-just-prediction-optimistic-test-time-world-model-adaptation-2610-35636Chiosaen2026-10-27T09:16:52ZPlanning-aware updates help world models adapt to shiftshttps://chiosa.app/papers/adapt-for-planning-not-just-prediction-optimistic-test-time-world-model-adaptation-2610-35636?lang=frChiosafr2026-10-27T09:16:52ZDes mises à jour guidées par la planification aident les modèles du monde à s’adapter aux changementshttps://chiosa.app/papers/audit-where-hacks-pay-budgeted-strong-verifier-audits-against-reward-hacking-in-grpo-2610-35716Chiosaen2026-10-27T09:16:52ZTargeted audits nearly eliminate coding reward hackinghttps://chiosa.app/papers/audit-where-hacks-pay-budgeted-strong-verifier-audits-against-reward-hacking-in-grpo-2610-35716?lang=esChiosaes2026-10-27T09:16:52ZAuditorías dirigidas casi eliminan las trampas de recompensa en programaciónhttps://chiosa.app/papers/back-to-basics-a-skill-centric-approach-to-generalist-web-agents-2610-35823Chiosaen2026-10-27T09:16:52ZA planner helps browser agents use practiced web skillshttps://chiosa.app/papers/back-to-basics-a-skill-centric-approach-to-generalist-web-agents-2610-35823?lang=itChiosait2026-10-27T09:16:52ZUn planner aiuta gli agenti browser a usare abilità web esercitatehttps://chiosa.app/papers/back-to-basics-a-skill-centric-approach-to-generalist-web-agents-2610-35823?lang=esChiosaes2026-10-27T09:16:52ZUn planificador ayuda a los agentes a usar habilidades web practicadashttps://chiosa.app/papers/back-to-basics-a-skill-centric-approach-to-generalist-web-agents-2610-35823?lang=frChiosafr2026-10-27T09:16:52ZUn planificateur aide les agents de navigation à mobiliser leurs compétences webhttps://chiosa.app/papers/built-to-remember-shipped-to-forget-operational-contracts-failures-and-repairs-in-containerized-llm-agent-dep-2610-34620Chiosaen2026-10-27T09:16:52ZFourteen deployment guarantees expose risks to agent systemshttps://chiosa.app/papers/constrained-decoding-harmful-or-beneficial-revisiting-its-impact-on-llm-performance-2610-35328Chiosaen2026-10-27T09:16:52ZMatched tests find constrained decoding need not hurt accuracyhttps://chiosa.app/papers/constrained-decoding-harmful-or-beneficial-revisiting-its-impact-on-llm-performance-2610-35328?lang=itChiosait2026-10-27T09:16:52ZTest abbinati: il decoding vincolato non riduce necessariamente l’accuratezzahttps://chiosa.app/papers/constrained-decoding-harmful-or-beneficial-revisiting-its-impact-on-llm-performance-2610-35328?lang=esChiosaes2026-10-27T09:16:52ZLas pruebas controladas muestran que la decodificación restringida no tiene por qué reducir la precisiónhttps://chiosa.app/papers/constrained-decoding-harmful-or-beneficial-revisiting-its-impact-on-llm-performance-2610-35328?lang=frChiosafr2026-10-27T09:16:52ZDes tests contrôlés montrent que le décodage contraint ne nuit pas forcément à la précisionhttps://chiosa.app/papers/cosa-the-conflict-of-skills-attack-in-real-world-llm-agent-ecosystems-2610-35082Chiosaen2026-10-27T09:16:52ZChecked Skills can still combine into policy violationshttps://chiosa.app/papers/cosa-the-conflict-of-skills-attack-in-real-world-llm-agent-ecosystems-2610-35082?lang=itChiosait2026-10-27T09:16:52ZSkill verificate possono combinarsi e violare le regolehttps://chiosa.app/papers/cosa-the-conflict-of-skills-attack-in-real-world-llm-agent-ecosystems-2610-35082?lang=frChiosafr2026-10-27T09:16:52ZDes Skills vérifiées peuvent tout de même se combiner et entraîner des violations de règleshttps://chiosa.app/papers/discodemo-discovering-efficient-and-diverse-robot-demonstrations-for-imitation-learning-2610-32093Chiosaen2026-10-27T09:16:52ZDiverse robot demonstrations improve imitation learninghttps://chiosa.app/papers/dynaflow-learning-free-form-multi-agent-orchestration-2610-34467Chiosaen2026-10-27T09:16:52ZDynaFlow: Learning free-form multi-agent orchestrationhttps://chiosa.app/papers/efficient-probabilistic-weather-forecasting-at-quarter-degree-resolution-with-two-sided-transport-attention-2610-35545Chiosaen2026-10-27T09:16:52ZA weather model brings detail to global forecastshttps://chiosa.app/papers/efficient-probabilistic-weather-forecasting-at-quarter-degree-resolution-with-two-sided-transport-attention-2610-35545?lang=itChiosait2026-10-27T09:16:52ZUn modello meteorologico porta più dettaglio nelle previsioni globalihttps://chiosa.app/papers/efficient-probabilistic-weather-forecasting-at-quarter-degree-resolution-with-two-sided-transport-attention-2610-35545?lang=esChiosaes2026-10-27T09:16:52ZUn modelo meteorológico detalla los pronósticos globaleshttps://chiosa.app/papers/excise-locating-and-excising-distracting-sentences-in-medical-question-answering-2610-35558Chiosaen2026-10-27T09:16:52ZA sentence-level method reduces targeted distraction in medical QAhttps://chiosa.app/papers/excise-locating-and-excising-distracting-sentences-in-medical-question-answering-2610-35558?lang=itChiosait2026-10-27T09:16:52ZUn metodo per ridurre le distrazioni mirate nelle domande medichehttps://chiosa.app/papers/excise-locating-and-excising-distracting-sentences-in-medical-question-answering-2610-35558?lang=esChiosaes2026-10-27T09:16:52ZUn método por oración reduce las distracciones dirigidas en preguntas médicashttps://chiosa.app/papers/excise-locating-and-excising-distracting-sentences-in-medical-question-answering-2610-35558?lang=frChiosafr2026-10-27T09:16:52ZUne méthode phrase par phrase réduit les distractions ciblées en questions médicaleshttps://chiosa.app/papers/governing-enterprise-ai-traffic-at-scale-design-and-operation-of-a-policy-driven-multi-vendor-llm-gateway-on-2610-35509Chiosaen2026-10-27T09:16:52ZPayPal’s gateway brings multi-vendor AI access under shared ruleshttps://chiosa.app/papers/governing-enterprise-ai-traffic-at-scale-design-and-operation-of-a-policy-driven-multi-vendor-llm-gateway-on-2610-35509?lang=frChiosafr2026-10-27T09:16:52ZLa passerelle de PayPal unifie l’accès à plusieurs fournisseurs d’IA sous des règles communeshttps://chiosa.app/papers/hilo-towards-product-level-graceful-degradation-with-a-quality-aware-load-aware-control-framework-2610-34844Chiosaen2026-10-27T09:16:52ZHiLO coordinates service slowdowns to protect product qualityhttps://chiosa.app/papers/hilo-towards-product-level-graceful-degradation-with-a-quality-aware-load-aware-control-framework-2610-34844?lang=itChiosait2026-10-27T09:16:52ZHiLO coordina il rallentamento dei servizi per proteggere la qualitàhttps://chiosa.app/papers/hilo-towards-product-level-graceful-degradation-with-a-quality-aware-load-aware-control-framework-2610-34844?lang=esChiosaes2026-10-27T09:16:52ZHiLO coordina la reducción de actividad para proteger la calidad del productohttps://chiosa.app/papers/identities-not-pairs-distribution-free-false-accept-control-for-embedding-based-verification-2610-35277Chiosaen2026-10-27T09:16:52ZA trial-aware image test controls false matcheshttps://chiosa.app/papers/identities-not-pairs-distribution-free-false-accept-control-for-embedding-based-verification-2610-35277?lang=itChiosait2026-10-27T09:16:52ZUn test delle immagini tiene conto della prova per controllare i falsi matchhttps://chiosa.app/papers/imagetts-repurposing-pretrained-image-diffusion-transformers-for-speech-synthesis-2610-34807Chiosaen2026-10-27T09:16:52ZImage-generation models can help a speech system learn wordshttps://chiosa.app/papers/interactiveworld-a-verifiable-human-centric-environment-for-building-better-interactive-mllm-assistants-2610-33966Chiosaen2026-10-27T09:16:52ZA simulator tests assistants as advice changes eventshttps://chiosa.app/papers/octen-webgr-scaling-generative-retrieval-to-200-million-web-pages-2610-35256Chiosaen2026-10-27T09:16:52ZA language model ranks pages from search clickshttps://chiosa.app/papers/octen-webgr-scaling-generative-retrieval-to-200-million-web-pages-2610-35256?lang=esChiosaes2026-10-27T09:16:52ZUn modelo de lenguaje ordena páginas a partir de clics de búsquedahttps://chiosa.app/papers/octen-webgr-scaling-generative-retrieval-to-200-million-web-pages-2610-35256?lang=frChiosafr2026-10-27T09:16:52ZUn modèle de langage classe les pages à partir des clicshttps://chiosa.app/papers/one-megakernel-is-not-enough-predicting-where-to-cut-for-streaming-inference-2610-35336Chiosaen2026-10-27T09:16:52ZA planner finds where streaming models should fuse GPU workhttps://chiosa.app/papers/one-megakernel-is-not-enough-predicting-where-to-cut-for-streaming-inference-2610-35336?lang=esChiosaes2026-10-27T09:16:52ZUn planificador decide dónde fusionar el trabajo en la GPUhttps://chiosa.app/papers/silent-and-costly-characterizing-and-detecting-prompt-caching-mis-integration-in-llm-agents-2610-34817Chiosaen2026-10-27T09:16:52ZA code checker finds silent prompt-caching defectshttps://chiosa.app/papers/silent-and-costly-characterizing-and-detecting-prompt-caching-mis-integration-in-llm-agents-2610-34817?lang=itChiosait2026-10-27T09:16:52ZUn checker individua difetti silenziosi nel prompt cachinghttps://chiosa.app/papers/silent-and-costly-characterizing-and-detecting-prompt-caching-mis-integration-in-llm-agents-2610-34817?lang=esChiosaes2026-10-27T09:16:52ZUn analizador detecta defectos inadvertidos en la caché de instruccioneshttps://chiosa.app/papers/streamvibe-proactive-video-agents-for-evolving-user-intent-2610-34662Chiosaen2026-10-27T09:16:52ZA video agent tracks changing goals to time its helphttps://chiosa.app/papers/the-hidden-costs-of-forgetting-a-study-of-trajectory-deletion-in-software-engineering-agents-2610-35496Chiosaen2026-10-27T09:16:52ZDeleting agent history works best when recent tool results stayhttps://chiosa.app/papers/the-hidden-costs-of-forgetting-a-study-of-trajectory-deletion-in-software-engineering-agents-2610-35496?lang=esChiosaes2026-10-27T09:16:52ZLos agentes resuelven mejor las tareas si conservan los resultados recientes de las herramientashttps://chiosa.app/papers/the-price-of-scale-in-proactive-memory-recall-2610-34507Chiosaen2026-10-27T09:16:52ZWhy fixed thresholds can fail as memory stores growhttps://chiosa.app/papers/the-voice-on-the-line-2610-35272Chiosaen2026-10-27T09:16:52ZVoice isolation improves overlap, but still harms lone callershttps://chiosa.app/papers/the-voice-on-the-line-2610-35272?lang=itChiosait2026-10-27T09:16:52ZL’isolamento vocale migliora con voci sovrapposte, ma danneggia ancora chi parla da solohttps://chiosa.app/papers/trajectory2gym-decision-grounded-compositional-interaction-synthesis-for-tool-use-agents-2610-33568Chiosaen2026-10-27T09:16:52ZA method builds grounded tool-use training interactionshttps://chiosa.app/papers/what-effort-level-should-i-choose-on-the-economics-of-reasoning-effort-in-llm-agents-2610-34491Chiosaen2026-10-27T09:16:52ZHigher reasoning effort can miss cheaper successeshttps://chiosa.app/papers/world-in-context-interactive-video-world-modeling-with-in-context-conditioning-2610-33169Chiosaen2026-10-27T09:16:52ZExplicit view controls help video models separate motion and switch views