MCPO: Modality-Contrastive Preference Optimization for Multimodal Chain-of-Thought Compre…
Explorar
Noticias de IA
37834 elementos — filtrados, clasificados y sin duplicados
Building a research-software catalog with a coding agent: from hackathon prototype to pub…
Refuse without Refusal: A Structural Analysis of Safety-Tuning Responses for Reducing Fal…
Tracing Audio Grounding and Answer Selection in Audio LLMs
Beyond Code Generation: Reliability, Verification, and Cost Economics in the Agentic Soft…
Enhancing Multimodal Emotion Recognition via Multi-Feature Encoding and Attention-Based F…
When Do Internal Probes Beat Reading the Answer? Miscalibrated Readouts and Behavior-Conc…
Persistent Teacher Anchoring for Tool-Using Agents
Dynamic Heterogeneous Graph Representation Learning: A Survey
Atlas: Optimizing Deployment of Compound AI Workflows on Heterogeneous Clusters
Hakken: Predicting future discoveries to fill the gaps in today's knowledge
Repeat-After-Me: Black-Box Adaptive Visual Prompt Injection
Patterns of Priming in Production: Lexical, Semantic and Structural Alignment in Language…
Dynamic Adaptation of the LLM Context for Generating Routines with Coupled Semantics
Cultural Misalignment in Large Language Models: Detection, Measurement, and Mitigation Th…
Training-Free Halving of Activated Experts in Fine-Grained Mixture-of-Experts Models
ACE: Adaptive Calibration-Free Expert Skipping for MoE-based LLMs
Reflection-aware Generative Novel View Synthesis
Scalable Context Orchestration for Serving LLMs Over Voice
Continual Graph Memory for Adaptive Recommendation under Intent Drift
AutoLR: Automating the Path from Research to Launch Review in Industrial Recommender Syst…
Moral Competence Before Moral Content: Why LLM Agents Lack the Prerequisites for Coherent…
A Roadmap for MEG Foundation Models
Australia’s AI Boom Risks Demand Surge and Higher Rates, BE Says
Claude proves Fermat 🧮, automated AI researcher 🔬, Z1 efficiency chip ⚡
Supporting independent journalism in Ukraine
Seattle Times and Newsday sue OpenAI and Microsoft for infringement
It was a great week for innovation across the model ecosystem. We’re bringing these new m…
Authors push back as publishers and agents make claims on Anthropic settlement
OpenAI says it reached its goal of creating an automated research intern