Publications

Periodic Weak Spots: Phase Sensitivity from Chunked KV-Cache Compression

HarnessDev: Can LLMs Create and Evolve Their Own Agent Harness?

S3Gym: Can LLMs Turn Self-Testing and Self-Judging into Self-Improvement?

Aspire: Can Models Self-Evolve from Vague Goals?

Chain-of-Experience for Continual LLM Improvement

Harness-IF: Evaluating Instruction Following Across Instruction Surfaces in Coding Agents

GST-Bench: Can VLMs Develop Global Spatial Awareness from Video?

Sample-Efficient Learning from Agent Experience

EdgeBench: Unveiling Scaling Laws of Learning from Real-World Environments

MSQA: A Natively Sourced Multilingual and Multicultural SimpleQA Benchmark

Seed2.0 Model Card: Towards Intelligence Frontier for Real-World Complexity

A Unified Generative Framework for Scalable Chemical Reaction Network Exploration

Analytic Nonadiabatic Derivative Couplings Using Noncollinear Spin-Flip TDDFT

Towards stable and accurate electron dynamics via neural network based time-dependent variational Monte Carlo

MetaPoint: Unlocking Precise Spatial Control in Agentic Visual Generation

Task-Focused Memorization for Multimodal Agents

More Expressive Feedforward Layers: Part I. Token-Adaptive Mixing of Activations

Negligible in Size, Significant in Effect: On Scale Vectors in Large Language Models