Scaling the Horizon, Not the Parameters: Reaching Trillion-Parameter Performance with a 35B Agent
Paper • 2606.30616 • Published • 86
Qihan Ren
jasonrqh
AI & ML interests
XAI, LLM reasoning & safety, Coding agent
Recent Activity
upvoted a paper 9 days ago
The Verification Horizon: No Silver Bullet for Coding Agent RewardsOrganizations
upvoted a paper 5 days ago
upvoted a paper 9 days ago
upvoted a paper 16 days ago
upvoted 2
papers 23 days ago
upvoted 3
papers about 1 month ago
COLLEAGUE.SKILL: Automated AI Skill Generation via Expert Knowledge Distillation
Paper • 2605.31264 • Published • 123
The MiniMax-M2 Series: Mini Activations Unleashing Max Real-World Intelligence
Paper • 2605.26494 • Published • 41
AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security
Paper • 2605.29801 • Published • 144
upvoted 3
papers about 2 months ago
MMSkills: Towards Multimodal Skills for General Visual Agents
Paper • 2605.13527 • Published • 122
OpenSeeker-v2: Pushing the Limits of Search Agents with Informative and High-Difficulty Trajectories
Paper • 2605.04036 • Published • 72
ARIS: Autonomous Research via Adversarial Multi-Agent Collaboration
Paper • 2605.03042 • Published • 141
upvoted 3
papers 3 months ago
Seedance 2.0: Advancing Video Generation for World Complexity
Paper • 2604.14148 • Published • 168
Turing Test on Screen: A Benchmark for Mobile GUI Agent Humanization
Paper • 2604.09574 • Published • 30
Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe
Paper • 2604.13016 • Published • 113
upvoted an article 3 months ago
Article
Red Teaming with RL: Exploiting Tinker API for Harmful RL on 235B Model
georgefen
• • 19 upvoted 5
papers 3 months ago
Rethinking Generalization in Reasoning SFT: A Conditional Analysis on Optimization, Data, and Model Capability
Paper • 2604.06628 • Published • 329
Why Does Self-Distillation (Sometimes) Degrade the Reasoning Capability of LLMs?
Paper • 2603.24472 • Published • 57
On the Role of Reasoning Patterns in the Generalization Discrepancy of Long Chain-of-Thought Supervised Fine-Tuning
Paper • 2604.01702 • Published • 3
SIM1: Physics-Aligned Simulator as Zero-Shot Data Scaler in Deformable Worlds
Paper • 2604.08544 • Published • 16
ATBench: A Diverse and Realistic Trajectory Benchmark for Long-Horizon Agent Safety
Paper • 2604.02022 • Published • 15