01

ConvictionBench

Evidence-sensitive belief revision in language models

When challenged, a language model has to decide whether to keep its answer, revise it, or say the evidence is insufficient. ConvictionBench tests how that decision changes with the evidence and with the pressure to agree.

02

Evidence and social pressure

ConvictionBench v0.1 contains 26,478 generated episodes across arithmetic, ledgers, logic, and small rule-based worlds. A deterministic solver determines the expected answer and decision.

Matched episodes vary the wording of a challenge, the quality of its evidence, and the channel the evidence arrives through. The structural split holds out whole problem types. A separate natural-history evaluation challenges the model’s own first answer.

In a control using identical record text, Anchor adopted the record 62% of the time from a tool message and 58% of the time from a plain user message, compared with 8% when the user added the forger’s wording. That result shows how much the measured resistance depended on the framing.

Same record, different framing

How often Conviction Anchor 1.7B adopted an identical record text, by the channel and wording it arrived in.

  • Tool message62%
  • Plain user message58%
  • User + forger’s wording8%

The tasks use short synthetic dialogues and a fixed response format. The results describe behavior under those conditions and give limited evidence about real-world authentication or manipulation.

03

Laplace Conviction Series

Three experimental Qwen3-1.7B LoRA adapters test different ways to train for belief revision.

Anchor often kept its own wrong answers after valid corrections in natural history. Vector had the highest held-out final accuracy of the three in the original setup, with weaker forged-record resistance than the base model on the same split. Its accuracy advantage over Anchor was sensitive to the seed and prompt. Meridian resisted more forged records than the other adapters on held-out types and had lower final accuracy than Vector.

Held-out problem types

Point estimates with 95% cluster-bootstrap intervals. Scripted structural split, greedy decoding, thinking disabled. Axes run from 0.2 to 0.8.

  • Anchor · SFT
  • Vector · GRPO
  • Meridian · light SFT, then GRPO
  • Qwen3-1.7B base

Final accuracy · higher is better

  • Anchor
  • Vector
  • Meridian
  • Qwen3 base

Pressure flips · lower is better

  • Anchor
  • Vector
  • Meridian
  • Qwen3 base

Forged-record resistance · higher is better

  • Anchor
  • Vector
  • Meridian
  • Qwen3 base
Results on held-out problem types
Model Training Final accuracy Pressure flips Forged-record resistance
Conviction Anchor 1.7B Supervised fine-tuning 0.580 0.389 0.531
Conviction Vector 1.7B GRPO from the base model 0.644 0.348 0.315
Conviction Meridian 1.7B Light supervised fine-tuning, then GRPO 0.560 0.375 0.578

Point estimates from the scripted structural split, using greedy decoding with thinking disabled. Lower pressure-flip rates are better. The Qwen3-1.7B base scored 0.424, 0.536, and 0.340 on the same measures. Intervals for every model are on the leaderboard.

04

Paper and reproducibility

ConvictionBench: Evidence-Sensitive Belief Revision in Language Models

The Laplace Research edition credits Arya Somu and Bruce Nshuti Hirwa. Arya Somu first released the frozen v0.1 benchmark on . Laplace Research became its publisher and maintainer on .

The code repository provides the benchmark generator, evaluation harness, experiment logs, and frozen file hashes. Code and adapters are Apache-2.0. The dataset is CC BY 4.0 and subject to its Hugging Face access agreement.

Cite the Laplace Research edition

@techreport{somu2026convictionbench-laplace,
  title       = {{ConvictionBench}: Evidence-Sensitive Belief Revision in Language Models},
  author      = {Somu, Arya and Nshuti Hirwa, Bruce},
  institution = {Laplace Research},
  year        = {2026},
  month       = nov,
  type        = {Preprint},
  note        = {ConvictionBench v0.1, Laplace Research edition},
  url         = {https://laplaceresearch.org/convictionbench/paper.pdf}
}