https://chiosa.app/papers/curvetq-rotation-free-trellis-quantization-of-llm-weights-via-curvature-weighted-search-2610-09212Chiosaen2026-10-11T00:15:48ZCurvature-guided compression avoids rotating model weightshttps://chiosa.app/papers/grace-generation-aware-latent-compression-for-efficient-video-generation-2610-10524Chiosaen2026-10-10T14:41:28ZVideo compression cuts generation time while preserving Wan2.1 scoreshttps://chiosa.app/papers/inverting-multi-vector-visual-document-indices-2610-09920Chiosaen2026-10-10T14:41:28ZStored page vectors can reveal the document they representhttps://chiosa.app/papers/real-time-joint-audio-video-generation-by-parallel-adapter-composition-2610-10343Chiosaen2026-10-10T14:41:28ZTwo adapters let audio–video generation stream in a few stepshttps://chiosa.app/papers/scalable-patch-level-self-supervised-learning-2610-10013Chiosaen2026-10-10T14:41:28ZPatch-level visual learning scales with a simpler objectivehttps://chiosa.app/papers/alaya-evoke-from-linear-scaling-supervision-to-endless-world-2608-13546Chiosaen2026-10-10T14:28:49ZEvoke uses bounded scene memory for long-video sessionshttps://chiosa.app/papers/chronos-enables-code-agents-to-reason-over-software-evolution-2610-11578Chiosaen2026-10-10T14:28:49ZChronos helps code agents search connected project historyhttps://chiosa.app/papers/dial-opd-learning-more-from-fewer-tokens-in-on-policy-distillation-2610-11659Chiosaen2026-10-10T14:28:49ZChoosing fewer, better tokens improves distillationhttps://chiosa.app/papers/diffuplex-accelerating-full-duplex-spoken-dialog-models-via-rolling-masked-diffusion-2610-12214Chiosaen2026-10-10T14:28:49ZDiffuPlex lets spoken-dialog models predict ahead and revise draftshttps://chiosa.app/papers/early-signatures-of-memorization-in-diffusion-models-via-basin-geometry-and-cyclic-denoising-2610-11670Chiosaen2026-10-10T14:28:49ZDiffusion models can reveal memorized images before copieshttps://chiosa.app/papers/easy-to-anticipate-hard-to-compute-boundary-dependence-finds-the-computed-outputs-that-entropy-patching-misse-2610-11790Chiosaen2026-10-10T14:28:49ZByte models miss results that need calculationhttps://chiosa.app/papers/from-public-posts-to-ai-search-citations-measuring-the-fragility-of-ai-search-2610-11932Chiosaen2026-10-10T14:28:49ZOrdinary web posts can enter AI-search answershttps://chiosa.app/papers/glio2-a-gpu-parallelized-tightly-coupled-lidar-inertial-gnss-system-for-robust-and-real-time-global-localizat-2610-12411Chiosaen2026-10-10T14:28:49ZGLIO2 completes a route with weak LiDAR geometryhttps://chiosa.app/papers/gpu-accelerated-computation-of-persistent-homology-for-topological-analysis-of-image-data-2610-10959Chiosaen2026-10-10T14:28:49ZTopoGPU streams image topology calculations through GPUshttps://chiosa.app/papers/listen-to-reason-listen-with-experts-retrieve-over-a-graph-reason-with-llms-2610-10749Chiosaen2026-10-10T14:28:49ZA text-based method makes audio answers traceable to named soundshttps://chiosa.app/papers/neural-network-verification-for-deep-joint-source-channel-coding-2610-11994Chiosaen2026-10-10T14:28:49ZA verifier bounds image errors across wireless channelshttps://chiosa.app/papers/omni-diffusion-distill-few-step-distillation-of-unified-multimodal-diffusion-large-language-models-2610-10990Chiosaen2026-10-10T14:28:49ZOne model learns faster image generation and understandinghttps://chiosa.app/papers/on-the-risks-of-using-llm-generated-tests-for-regression-testing-2610-11835Chiosaen2026-10-10T14:28:49ZGenerated tests can preserve bugs as expected behaviorhttps://chiosa.app/papers/one-skill-too-many-how-co-installed-skills-conflict-in-coding-agents-2610-11647Chiosaen2026-10-10T14:28:49ZSimilar coding skills can hide the behavior users installedhttps://chiosa.app/papers/rehearse-everything-remember-nothing-attic-kv-rehearses-what-will-be-read-2610-12133Chiosaen2026-10-10T14:28:49ZA cache compressor rehearses likely answers before questions arrivehttps://chiosa.app/papers/right-screen-wrong-transition-world-models-as-verifiers-for-gui-agents-2610-11942Chiosaen2026-10-10T14:28:49ZA world model checks whether GUI actions lead as expectedhttps://chiosa.app/papers/should-your-database-systems-use-hardware-assisted-memory-safety-extensions-in-production-2610-11525Chiosaen2026-10-10T14:28:49ZDatabase memory checks trade lower cost for stronger protectionhttps://chiosa.app/papers/skill-constellations-tracing-the-supply-chain-of-agent-skills-on-github-2610-11169Chiosaen2026-10-10T14:28:49ZGit histories reveal how agent skills spreadhttps://chiosa.app/papers/sr-tta-spatial-redundancy-test-time-adaptation-for-interference-robust-respiration-sensing-2610-11755Chiosaen2026-10-10T14:28:49ZAntenna agreement helps recover breathing under motion interferencehttps://chiosa.app/papers/stop-my-dancing-understanding-detecting-and-attributing-motion-aware-deepfake-videos-2610-11496Chiosaen2026-10-10T14:28:49ZMotion fakes leave clues MoDA uses to detect and tracehttps://chiosa.app/papers/the-lattice-of-transition-laws-2610-11216Chiosaen2026-10-10T14:28:49ZA lattice method predicts better decoding schedules before generationhttps://chiosa.app/papers/the-learner-who-does-not-learn-when-optimizing-a-pedagogical-metric-degrades-llm-tutoring-2610-12125Chiosaen2026-10-10T14:28:49ZA tutor's higher fit score came with lower educator ratingshttps://chiosa.app/papers/tokenrouter-efficient-serving-system-for-token-level-llm-routing-2610-12242Chiosaen2026-10-10T14:28:49ZTokenRouter helps models hand off requests without restarting workhttps://chiosa.app/papers/viola-learning-generalist-humanoid-control-policies-from-human-data-2610-12435Chiosaen2026-10-10T14:28:49ZHuman motion helps humanoid robots follow instructionshttps://chiosa.app/papers/when-do-we-need-on-policy-distillation-distilling-on-offline-student-rollouts-is-often-better-2610-11291Chiosaen2026-10-10T14:28:49ZReusing a student's first answers can improve distillation