Pith. sign in

Are reasoning llms robust to interventions on their chain-of-thought? InThe Fourteenth International Conference on Learning Representations (ICLR)

2 Pith papers cite this work. Polarity classification is still indexing.

2 Pith papers citing it

citation-role summary

background 1

citation-polarity summary

fields

cs.CL 1 cs.LG 1

years

2026 2

verdicts

CONDITIONAL 2

roles

background 1

polarities

background 1

representative citing papers

Robust Reasoning Benchmark

cs.LG · 2026-03-26 · conditional · novelty 6.0 · 2 refs

A 13-way text-scrambling benchmark makes open-weight LLMs drop up to 54% average accuracy, and a multi-problem prompt makes their accuracy on the last question decay.

citing papers explorer

Showing 2 of 2 citing papers.

  • Evaluation Awareness in Language Models Has Limited Effect on Behaviour cs.CL · 2026-05-07 · conditional · none · ref 25

    Verbalised evaluation awareness in large reasoning models has only small effects on their outputs across safety and alignment tests.

  • Robust Reasoning Benchmark cs.LG · 2026-03-26 · conditional · none · ref 53 · 2 links

    A 13-way text-scrambling benchmark makes open-weight LLMs drop up to 54% average accuracy, and a multi-problem prompt makes their accuracy on the last question decay.