This work asks whether a model can rewrite its own weights to make itself better, and finds that it can, by its own reasoning, with no gradients and no training.
Jia-Jun Fan, Neil He, Chang-Yi Yang et al.· 0 citations
An unsupervised multi-source mutation preference distillation framework that learns from relative mutation preferences while explicitly modeling cross-source disagreement is proposed that achieves the best overall performance among the evaluated zero-shot baselines and naive multi-source fusion strategies on ProteinGym...
Rui-Han Guo, Chao-Ran Cheng, Zhang-Han Ni et al.· 0 citations
Most unified multimodal models treat images and text asymmetrically: a pretrained language model generates text, an attached diffusion head generates images, and outputs that contain both are produced in a fixed order, so the first output cannot adapt to the second. We ask whether both modalities can instead be generat...
Jia-Han Li, Huang-Yuan Su, Sheng-Bang Tong et al.· 0 citations
Results identify reward-preserving training structure as a practical route to lower per-step cost in short-window RLVR, and single-problem evaluations provide supporting evidence of retained reasoning quality.
Bang Yang, Hong-Bo Ma, Rui-Han Guo et al.· 0 citations
How much can test-time computation improve a language model, and at what cost? Test-time scaling is widely proposed as a substitute for larger models, but existing comparisons mostly evaluate one domain at a time and rarely charge selection to the budget. We introduce SELF-POT, a benchmark and evaluation framework that...
Bang Yang, Jing-Yuan Li, Jia-Jun Fan et al.· 0 citations
Scaling reasoning typically spends more compute on reinforcement learning (RL) or on inference. We show that a completed RL training history can yield policies stronger than the checkpoints visited by its optimizer. We call this policy-space scaling: expanding the deployable policy set accessible from a fixed RL histor...
Bang Yang, Jia-Jun Fan, Hong-Bo Ma et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.