α
Research
Alpha Leak
Conferences
Topics
Top Authors
Rankings
Browse All
EN
中
Home
/
Authors
/
Tom Bewley
Tom Bewley
5
papers
36
total citations
papers (5)
Non-Markovian Reward Modelling from Trajectory Labels via Interpretable Multiple Instance Learning
NEURIPS 2022
arXiv
19
citations
To Steer or Not to Steer? Mechanistic Error Reduction with Abstention for Language Models
ICML 2025
arXiv
6
citations
Interpreting Language Reward Models via Contrastive Explanations
ICLR 2025
arXiv
5
citations
Counterfactual Metarules for Local and Global Recourse
ICML 2024
arXiv
4
citations
Representation Consistency for Accurate and Coherent LLM Answer Aggregation
NEURIPS 2025
arXiv
2
citations