Selecting Diverse SFT Traces Improves Post-RL Generalization
Paper • 2609.33780 • Published • 38
Jan Novák
jan-novak
·
AI & ML interests
None yet
Recent Activity
liked a dataset 18 days ago
ukisai/Qwen3.8-27B-multi-turn-agent-sft liked a model 18 days ago
Tongyi-MAI/Z-Image-Turbo liked a model 18 days ago
bottlecapai/ThinkingCap-Qwen3.8-27BOrganizations
None yet
upvoted 5
papers 18 days ago
Preference-Guided Adaptation for Open-Vocabulary Semantic Segmentation via Prompt Disagreement
Paper • 2609.34528 • Published • 7
CaptchaArena: A Large-Scale, Fine-Grained Dataset for Training Computer-Use Agents on Interactive CAPTCHAs
Paper • 2609.31957 • Published • 7
What Makes Recurrence Effective in Looped Language Models?
Paper • 2609.36636 • Published • 11
Scaling Properties of Same-Family On-Policy Distillation
Paper • 2609.32722 • Published • 327