arXiv ·
RAG-Stress: Probing the Limits of Evidence Reliance in Retrieval-Augmented Generation
Public paper from arXiv. Open the source to read the full record.
Home / Browse
Models & research
Examine models, papers and evaluations with their methods and limitations.
802 public stories
arXiv ·
Gated Memory: Admission-Controlled Memory Formation for Conversational AI
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Phonological Interference in Multilingual Speech Models
Public paper from arXiv. Open the source to read the full record.
arXiv ·
REMORY: Learning Residual Memory for Context Compaction
Public paper from arXiv. Open the source to read the full record.
arXiv ·
When Do We Need On-Policy Distillation? Distilling on Offline Student Rollouts Is Often Better
Public paper from arXiv. Open the source to read the full record.
arXiv ·
BeliefScope: Diagnosing Evidence-Driven Revision and Pressure-Induced Shifts in Large Language Models
Public paper from arXiv. Open the source to read the full record.
arXiv ·
An Explainable Header-Centric Framework for Large-Scale Semantic Table Interpretation and Data Quality Assessment
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Synthesis Through Simulation: Generating Coherent Enterprise Data via Scalable Agent-System Interaction
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Agent-Controlled Forgetting for Tool-Using Agents: Reversible Context Curation in Practice
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Verification and Self-Improvement in Agentic AI: Foundations and Limits
Public paper from arXiv. Open the source to read the full record.
arXiv ·
The Harness as the Only Mutable Surface: Compliance-Bounded Self-Evolution of LLM Agents in Credit Pipelines, with a Measured Admission Gate
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Speaking the Navigator's Language: Trajectory-Grounded Instruction Translation for Frozen Aerial VLN Agents
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Plan-and-Patch: Diffusion Language Models for Agentic Planning
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Whose Ground Truth? Embracing Ambiguity in Human-Centered AI
Public paper from arXiv. Open the source to read the full record.
arXiv ·
On the Clock: Towards Punctual and Productive Time-Budgeted AI Agents
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Self-Supervised Keyframe Discovery for Horizon-Invariant Behavior Cloning
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Reading the Room: Foundations, Design, and Challenges of Normative Competence in LLMs
Public paper from arXiv. Open the source to read the full record.
arXiv ·
StoreBench: A Live-Commerce Environment for Evaluating and Training Autonomous Operator Agents
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Learning How to Search for Plans with Exponentially Less Space
Public paper from arXiv. Open the source to read the full record.
arXiv ·
How Narrative Wrapping Affects LLM Refusal: A Cross-Language Benchmark and Defense
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Curating Always-Loaded Context for LLM Agents: A Capacitated Assortment Model with Censored Feedback
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Distillation for Incrimination and Distillation for Capabilities
Public paper from arXiv. Open the source to read the full record.
arXiv ·
AgentHorizon: Evaluating Agentic Judges for Long-Horizon Computer-Use Tasks
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Beyond Imitation: A Framework and Benchmark for LLM-Assisted Peer Review
Public paper from arXiv. Open the source to read the full record.
arXiv ·
OpenProblemBench: Benchmarking AI on Open Problems in the Foundational Theoretical Sciences
Public paper from arXiv. Open the source to read the full record.
arXiv ·
When Interfaces Speak: Data-Aware Generative UI Harness for Active Interaction
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Balancing Reference Guidance and Free Generation in Trajectory Rollouts for Reasoning RL
Public paper from arXiv. Open the source to read the full record.
arXiv ·
GameCommBench: A Unified Benchmark and Type-Aware Evaluation for AI-Generated Game Commentary
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Social Pain Disrupts Emotion-Action Brain-State Dynamics in Adolescents with Non-Suicidal Self-Injury
Public paper from arXiv. Open the source to read the full record.
arXiv ·
What to Admit and How to Present: Governing Persistent Memory in LLM Agents
Public paper from arXiv. Open the source to read the full record.