arXiv ·
FlashMap: Flow Map Speculative Decoding
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 ·
Ask Less, Ask Better: Allocating User Attention in Proactive LLM Agents with Interruption Budgets
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Toward a Universal Local Speech Feature Extractor Through Distillation
Public paper from arXiv. Open the source to read the full record.
arXiv ·
When Opposite Premises Translate Identically: A Source-Conditioned Diagnostic for Multilingual Commonsense Items
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Which Long Documents Matter? A Controlled Study of Data Selection for Context Extension
Public paper from arXiv. Open the source to read the full record.
arXiv ·
GardenBench: A lightweight, daily evaluation of LLM capabilities
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Doing Less with More: A First-Principles Exploration of the Suitability of Agentic Computing over Alternative Architectural Choices
Public paper from arXiv. Open the source to read the full record.
arXiv ·
A Controlled LLM Workflow for Evidence-Grounded Diagnosis of Risk Score Differences
Public paper from arXiv. Open the source to read the full record.
arXiv ·
What Social Media Teaches Us About Accumulative AI Risks
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Beyond Explicit Intent: Discovering and Shaping Artistic Intent through Human-AI Co-Creation
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Unified Multiparadigm Fusion Intelligence for Rare Disease Diagnosis
Public paper from arXiv. Open the source to read the full record.
arXiv ·
LogicVal:A Logic-Formalized Benchmark for Constraint Satisfaction in Text-to-Image Generation
Public paper from arXiv. Open the source to read the full record.
arXiv ·
World of SWECraft: Agent Collaboration at Scale via Simulated Software Worlds
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Persistent Memory and Iterative Improvement in an LLM Agent for Chinese-Character Drawing
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Cache2LoRA: Understanding and Improving Context Internalization
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Joint, Decoupled, or Agent: When Each Multimodal Alignment Paradigm Wins
Public paper from arXiv. Open the source to read the full record.
arXiv ·
A Conditional No-Hair Theorem for Transformers: How Language Models Forget Their Prompt
Public paper from arXiv. Open the source to read the full record.
arXiv ·
TorchTPU - A Hybrid Eager/Compiled Architecture for Native PyTorch Execution on TPU
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Self-Improving Video-to-Action Models
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Agents' Last Doubt: Knowing When to Act and When to Step Back in Tool-Using Agents
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Distillation, Self-Training, and the Fragility of Learned Chemical Reasoning
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Training-Free Prompt Selection for Efficient Reinforcement Learning With Partial Rollouts
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Probing Ambiguity: Adversarial Intent Resolution for Robust Text-to-SQL
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Cortex: A Fixed-Point Theory of Governed Coding Agents - Technical Report
Public paper from arXiv. Open the source to read the full record.
arXiv ·
AM-Bench: A Unified Taxonomy and Evaluation Suite for Agentic Misalignment
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Can Compact Open-Weight Large Language Models Recover Mobility Events from Incomplete Trajectories? A Controlled Evaluation Using Air Trips
Public paper from arXiv. Open the source to read the full record.
arXiv ·
From SEC Filings to Financial Signals: Evaluating LLMs on Stock Market Reactions and Analyst Forecasts
Public paper from arXiv. Open the source to read the full record.
arXiv ·
SteerSearch: Interactive Search Reasoning with Mid-Response Steering
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Not Every User Says It All: Training Tool-Using Agents Beyond Ideal Simulated Users
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Who Speaks, Who Authorizes, Who Acts? Understanding Trust Boundaries in Multi-Agent LLMs
Public paper from arXiv. Open the source to read the full record.