AI ethics evidence index Search
Source-linked daily record
Evidence record 1437 · automatically gathered

A Pre-Registered Causal Partition of Self-Consistency Elicitation and Reward Design in RLVR

Reinforcement learning from verifiable rewards (RLVR) improves reasoning even when the reward signal is spurious -- assigning credit to the group-plurality answer rather than a ground-truth verifier. Practitioners commonly interpret naive = acc(TRUE) - acc(RANDOM) as the reward-design effect. We prove this estimand is systematically biased: it conflates self-consistency elicitation (sharpening the policy toward its modal answer via majority pseudo-reward) with genuine reward-design signal. Using

Record details

Published: 4 June 2026
Source: arXiv
Category: Research
Topics: Bias & fairness · Regulation
Retrieved: 14 July 2026

source-onlyevidence status

These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.

Shared topic: Bias & fairness · Shared topic: Regulation

BiasGRPO: Stabilizing Bias Mitigation in High-Variance Reward Landscapes via Group-Relative Policy Optimization

arXiv · 3 June 2026

Shared topic: Bias & fairness · Shared topic: Regulation

The Fair Lending Model: How the Longest-Running Algorithmic Fairness Programs Work in Practice

arXiv · 1 June 2026

Shared topic: Bias & fairness · Shared topic: Regulation

Testing the Black Box: Structural Barriers to Independent Evaluation of Consumer-Facing Health LLMs

arXiv · 7 June 2026

Shared topic: Bias & fairness · Shared topic: Regulation

Reinforcement Learning for Flow-Matching Policies with Density Transport

arXiv · 7 June 2026

Shared topic: Bias & fairness · Shared topic: Regulation

Bridging Expert Knowledge and Automated Feature Engineering via Self-Evolution

arXiv · 7 June 2026

Shared topic: Bias & fairness · Shared topic: Regulation

AGORA: Can Deliberation and Governance Gates Absorb Participation Bias in Transit Planning?

arXiv · 31 May 2026

How to cite this record

ethics.ai (4 June 2026), “A Pre-Registered Causal Partition of Self-Consistency Elicitation and Reward Design in RLVR,” evidence record 1437, https://ethics.ai/record/1437 (originally published by arXiv).

JSON

Use and limitations

This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.