New Model Releases
A
Showing 151–180 of 276
-
Efficient Agentic Audio-Visual Question Answering through Toolset Pruning and Modality Consolidation
-
HEDGE: Gated Hypergraph Memory for Scalable Multi-Agent Debate
-
Contextual Embeddings Beyond the Gold Chunk
-
Autonomous Quantum Measurement with Agentic AI
-
Discovering Jailbreaks via Curriculum of Models
-
CopyLoop: Fast Looped Transformer Decoding via Context Copying and Shallow Screening
-
From Grind to Grasp
-
inkGround: Benchmarking Fine-grained Visual Grounding Capabilities of Vision-Language Models for Cursive Handwriting in the Wild
-
Learning Scientific Trajectories to Predict Future Biomedical Discoveries
-
Latent Decision Flow Rewards for Clinical Guideline Reasoning
-
Geometry Matters in Packed 3D Attention
-
Beyond fabrication: Misattributed numbers in papers written by AI research agents
-
Quantum-Inspired Pairwise Ranking Aggregation via Active LLM Feedback
-
Weaken to Strengthen: Weakening-Induced On-Policy Self-Distillation for Large Reasoning Models
-
AskStream: When and What to Ask Proactively under Uncertainty in Streaming Video Understanding
-
Explore Where It Matters: Targeted Exploration for Efficient Rover Adaptation
-
オラクル、社員が業務アプリをAIで自動作成、Oracle Cloud上で共有できる「Oracle App Composer」発表。Oracle AI World 2026Oracle unveils App Composer for AI-built apps on Oracle CloudAt Oracle AI World 2026 in Las Vegas, Oracle announced Oracle App Composer, which lets employees generate business apps with AI, deploy them on Oracle Cloud and share them there. Details beyond the announcement were not in the source excerpt.
-
The October inflection point for vibe game developmentWillison: October 2026 is an inflection point for vibe-coding gamesSimon Willison argues that October 2026 marks a new inflection point, this time for vibe coding computer games, following November 2025's shift for coding agents. He credits improvements in Claude Opus 5.5 and showcases games he built, such as Tree Appreciator and Emberwick.
-
Cross-Primitive Consistency in Jev
-
Think Before You Act: Decision Gate for LLM Agents
-
How Deterministic SAST and AI Security Agents Divide the Work: Coverage, Harness Policy and Cost in Snyk VulnBench 2.0
-
Rethinking LLM Jailbreak Evaluation: Continuous Risk Landscapes Beyond Binary Outcomes
-
Do Audio LLMs HEAR Pain? Tracing Acoustic Evidence from Speech to Language Reasoning
-
A-RSI: Adaptive Task Refresh for Zero-Data Multimodal Self-Improvement
-
Persona Learning from Feedback for Personalized AI Assistants
-
From Leniency to Suspicion: How an Open Post-Training Recipe Changes Whom LLM Fact-Checkers Believe
-
Value Probes: Learned Value Functions for Beam-Search Red Teaming
-
RoleFUZZER: User-Behaviour Fuzzing for Role-Specification Adherence in Large Language Models
-
DecisionUnlearn-Bench: Unlearning Probabilistic Decisions under Changing Schemas
-
SpaceAct: Benchmarking 3D Spatial and Temporal Reasoning via Closed-Loop Action and Active Perception