Jian Yang

ModularRSI is proposed, a benchmark-disjoint, contrastive, and modular framework for generalizable harness evolution that contrasts successful and failed trajectories for the same task and aggregates evidence across tasks to identify recurring behavioral deficiencies.

Experiments on 4DMatch and 4DLoMatch show that both variants produce more accurate correspondences than the compared methods and improve downstream registration, with larger gains in low-overlap cases.

This work introduces AgentActionBench, a process-oriented benchmark for evaluating agent-based experiment reproduction across ML and AI4Science domains, and uses an MCP-based Action Recorder to capture agents' behaviour throughout the reproduction process and evaluates the resulting traces with paper-specific rubrics.

Aegis, trained with CyberFactory, outperforms existing models by 22.8 points in cybersecurity tasks, showcasing the power of agentic learning from real-world vulnerabilities.

OPD transfers reasoning skills rather than answers, revealing a complex interplay between teacher-student origins that can either enhance or hinder model capabilities.

Ditch the garment masks: a simple human mask is all you need to nail video virtual try-on in the wild.

Building agents that can reliably automate complex, multi-step workflows over local files and tools just got a whole lot easier.

Forget scaling laws: a carefully crafted data curriculum lets a 32B parameter model beat GPT-4.5 and Claude-4.5 on web navigation.

LLMs can now automatically evolve and optimize GPU kernels to beat hand-tuned and proprietary models like Gemini and Claude.

A new 32B code LLM trained specifically for industrial tasks crushes existing models on specialized domains like chip design and GPU kernel optimization, while remaining competitive on general coding benchmarks.

Code LLMs can achieve SOTA performance in agentic tasks by explicitly modeling the dynamic evolution of software logic across different training stages.

A 32B model trained entirely on synthetic data from InfTool outperforms models 10x larger on tool use, rivaling even Claude-Opus.