A controlled formal language task reveals fine-tuning outperforms in-context learning on in-distribution generalization but equals it on out-of-distribution, with ICL showing greater sensitivity to model size and tokenization.
Title resolution pending
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
BiasedTales-ML provides a parallel multilingual corpus of LLM-generated children's stories that reveals substantial cross-lingual differences in narrative attributes not captured by English-centric analyses.
Pref-CTRL trains a multi-objective value function on preferences to guide representation editing for LLM alignment, outperforming RE-Control on benchmarks with better out-of-domain generalization.
Align-Cultura introduces the CULTURAX dataset and shows that culturally fine-tuned LLMs improve joint HHH scores by 4-6%, cut cultural failures by 18%, and gain 10-12% efficiency with minimal leakage.
GroupDPO decouples group-wise preference optimization during backpropagation to cut peak memory while keeping the same gradients, allowing larger groups and consistent gains over single-pair DPO plus an NLL term on positives.
Coupled constraints on weight updates in a safety subspace and regularization of SAE-identified safety features preserve LLM refusal behaviors during fine-tuning better than weight-only or activation-only methods.
Both humans and LLMs trust content more when labeled human-authored than AI-generated, with LLMs showing denser attention to labels and higher uncertainty under AI labels, mirroring human heuristic patterns.
Controlled prompt interventions reveal strong affiliation bias in LLM peer reviews favoring top-ranked institutions, plus effects from seniority and publication history.
citing papers explorer
-
Preventing Safety Drift in Large Language Models via Coupled Weight and Activation Constraints
Coupled constraints on weight updates in a safety subspace and regularization of SAE-identified safety features preserve LLM refusal behaviors during fine-tuning better than weight-only or activation-only methods.
-
Label Effects: Shared Heuristic Reliance in Trust Assessment by Humans and LLM-as-a-Judge
Both humans and LLMs trust content more when labeled human-authored than AI-generated, with LLMs showing denser attention to labels and higher uncertainty under AI labels, mirroring human heuristic patterns.