arXiv:2608. 05423v1 Announce Type: cross Abstract: In generative training, a model produces an output and is penalised for its difference from an example.
Summary generated by The Flow from the publisher's feed. The full article lives at arXiv AI.
arXiv:2608. 15445v1 Announce Type: new Abstract: When a reward is correct on every training example yet consistent with more than one goal, a model can acquire an unintended one, a failure known as goal misgeneralization.
arXiv:2607. 09693v1 Announce Type: cross Abstract: Reinforcement learning (RL) has become the dominant paradigm for improving the reasoning capabilities of large language models, but it requires expensive training, curated data, and reward signals.
arXiv:2607. 09786v1 Announce Type: new Abstract: Length-penalized reinforcement learning can shorten chain-of-thought reasoning while hiding an influence that drives the model's answer.
arXiv:2606. 07856v1 Announce Type: new Abstract: When a language model trains on its own verified outputs, does it acquire capability beyond its base, or merely get better at expressing capability the base already had?
arXiv:2606. 00544v1 Announce Type: new Abstract: Modern language-model fine-tuning typically pairs each prompt with a single response, even though many prompts admit multiple valid completions.
arXiv:2603. 06957v2 Announce Type: replace-cross Abstract: We study post-training linear autoregressive models with outcome and process rewards.