arXiv:2606. 01202v1 Announce Type: new Abstract: Language models do not simply choose an answer at the output layer.
Summary generated by The Flow from the publisher's feed. The full article lives at arXiv AI.
arXiv:2608. 11694v1 Announce Type: cross Abstract: A benchmark score comes from a single phrasing of each problem.
arXiv:2605. 26795v2 Announce Type: replace Abstract: Chain-of-thought (CoT) prompting enhances large language model performance, yet what drives these gains remains unclear.
arXiv:2608. 14320v1 Announce Type: new Abstract: The anchoring effect is a cognitive bias in which an initial reference value shifts a later judgment toward itself.
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. 15733v1 Announce Type: cross Abstract: Instruction-tuned language models can answer the same causal-reasoning question differently after its English variable names are replaced by type-preserving placeholders, although the structural causal model and the gold answer are unchanged.
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.