Skip to content
← Back to explorer

Making Bias Non-Predictive: Training Robust LLM Reasoning via Reinforcement Learning

Qian Wang, Xuandong Zhao, Zirui Zhang, Zhanzhi Lou, Nuo Chen, Dawn Song, Bingsheng He · Feb 2, 2026 · Citations: 0

Open arXiv Find Implementation RSS feed Shortlist (0)

Data freshness

Extraction: Recent

Check recency before relying on this page for active eval decisions. Use stale pages as context and verify against current hub results.

Metadata refreshed

Apr 6, 2026, 12:42 PM

Recent

Extraction refreshed

Apr 6, 2026, 12:42 PM

Recent

Extraction source

Persisted extraction

Confidence unavailable

Abstract

Large language models (LLMs) increasingly serve as reasoners and automated evaluators, yet they remain susceptible to cognitive biases -- often altering their reasoning when faced with spurious prompt-level cues such as consensus claims or authority appeals.} Existing mitigations via prompting or supervised fine-tuning fail to generalize, as they modify surface behavior without changing the optimization objective that makes bias cues attractive. We propose \textbf{Epistemic Independence Training (EIT)}, a reinforcement learning framework grounded in a key principle: to learn independence, bias cues must be made non-predictive of reward. EIT operationalizes this through a balanced conflict strategy where bias signals are equally likely to support correct and incorrect answers, combined with a reward design that penalizes bias-following without rewarding bias agreement. Experiments on Qwen3-4B demonstrate that EIT improves both accuracy and robustness under adversarial biases, while preserving performance when bias aligns with truth. Notably, models trained only on bandwagon bias generalize to unseen bias types such as authority and distraction, indicating that EIT induces transferable epistemic independence rather than bias-specific heuristics. \revised{EIT further generalizes across benchmarks (MedQA, HellaSwag), model families (Llama-3.2-3B), and scales (Qwen3-8B), and outperforms distribution-shift methods (GroupDRO, IRM) without requiring environment labels.} Code and data are available at https://anonymous.4open.science/r/bias-mitigation-with-rl-BC47

Low-signal caution for protocol decisions

Use this page for context, then validate protocol choices against stronger HFEPX references before implementation decisions.

  • Structured extraction is still processing; current fields are metadata-first.
Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub

HFEPX Relevance Assessment

Signal extraction is still processing. This page currently shows metadata-first guidance until structured protocol fields are ready.

Best use

Background context only

Use if you need

A provisional background reference while structured extraction finishes.

Main weakness

Structured extraction is still processing; current fields are metadata-first.

Trust level

Provisional

Eval-Fit Score

Unavailable

Eval-fit score is unavailable until extraction completes.

Human Feedback Signal

Not explicit in abstract metadata

Evaluation Signal

Weak / implicit signal

HFEPX Fit

Provisional (processing)

Extraction confidence: Provisional

If you are doing eval pipeline work, start here:

Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub Tool-Use Eval Hub

Field Provenance & Confidence

Each key protocol field shows extraction state, confidence band, and data source so you can decide whether to trust it directly or validate from full text.

Human Feedback Types

provisional

None explicit

Confidence: Provisional Source: Persisted extraction inferred

No explicit feedback protocol extracted.

Evidence snippet: Large language models (LLMs) increasingly serve as reasoners and automated evaluators, yet they remain susceptible to cognitive biases -- often altering their reasoning when faced with spurious prompt-level cues such as consensus claims or authority appeals.} Existing mitigations via prompting or supervised fine-tuning fail to generalize, as they modify surface behavior without changing the optimization objective that makes bias cues attractive.

Evaluation Modes

provisional

Automatic metrics, Simulation environment

Confidence: Provisional Source: Persisted extraction inferred

Includes extracted eval setup.

Evidence snippet: Large language models (LLMs) increasingly serve as reasoners and automated evaluators, yet they remain susceptible to cognitive biases -- often altering their reasoning when faced with spurious prompt-level cues such as consensus claims or authority appeals.} Existing mitigations via prompting or supervised fine-tuning fail to generalize, as they modify surface behavior without changing the optimization objective that makes bias cues attractive.

Quality Controls

provisional

Not reported

Confidence: Provisional Source: Persisted extraction inferred

No explicit QC controls found.

Evidence snippet: Large language models (LLMs) increasingly serve as reasoners and automated evaluators, yet they remain susceptible to cognitive biases -- often altering their reasoning when faced with spurious prompt-level cues such as consensus claims or authority appeals.} Existing mitigations via prompting or supervised fine-tuning fail to generalize, as they modify surface behavior without changing the optimization objective that makes bias cues attractive.

Benchmarks / Datasets

provisional

Not extracted

Confidence: Provisional Source: Persisted extraction inferred

No benchmark anchors detected.

Evidence snippet: Large language models (LLMs) increasingly serve as reasoners and automated evaluators, yet they remain susceptible to cognitive biases -- often altering their reasoning when faced with spurious prompt-level cues such as consensus claims or authority appeals.} Existing mitigations via prompting or supervised fine-tuning fail to generalize, as they modify surface behavior without changing the optimization objective that makes bias cues attractive.

Reported Metrics

provisional

Accuracy, Agreement / Kappa

Confidence: Provisional Source: Persisted extraction inferred

Useful for evaluation criteria comparison.

Evidence snippet: Experiments on Qwen3-4B demonstrate that EIT improves both accuracy and robustness under adversarial biases, while preserving performance when bias aligns with truth.

Rater Population

provisional

Unknown

Confidence: Provisional Source: Persisted extraction inferred

Rater source not explicitly reported.

Evidence snippet: Large language models (LLMs) increasingly serve as reasoners and automated evaluators, yet they remain susceptible to cognitive biases -- often altering their reasoning when faced with spurious prompt-level cues such as consensus claims or authority appeals.} Existing mitigations via prompting or supervised fine-tuning fail to generalize, as they modify surface behavior without changing the optimization objective that makes bias cues attractive.

Human Data Lens

Structured extraction is still processing. Below are provisional signals inferred from abstract text only.

  • Potential human-data signal: No explicit human-data keywords detected.
  • Potential benchmark anchors: No benchmark names detected in abstract.
  • Abstract highlights: 3 key sentence(s) extracted below.

Evaluation Lens

Evaluation fields are currently inferred heuristically from abstract text.

  • Potential evaluation modes: Automatic metrics, Simulation environment
  • Potential metric signals: Accuracy, Agreement / Kappa
  • Confidence: Provisional (metadata-only fallback).

Research Brief

Deterministic synthesis

Large language models (LLMs) increasingly serve as reasoners and automated evaluators, yet they remain susceptible to cognitive biases -- often altering their reasoning when faced with spurious prompt-level cues such as consensus claims or authority appeals.} Existing mitigations via prompting or supervised fine-tuning fail to generalize, as they modify surface behavior without changing the optimization objective that makes bias cues attractive.

Generated Apr 6, 2026, 12:42 PM · Grounded in abstract + metadata only

Key Takeaways

  • Large language models (LLMs) increasingly serve as reasoners and automated evaluators, yet they remain susceptible to cognitive biases -- often altering their reasoning when faced with spurious prompt-level cues such as consensus claims or authority appeals.} Existing mitigations via prompting or supervised fine-tuning fail to generalize, as they modify surface behavior without changing the optimization objective that makes bias cues attractive.
  • We propose \textbf{Epistemic Independence Training (EIT)}, a reinforcement learning framework grounded in a key principle: to learn independence, bias cues must be made non-predictive of reward.
  • EIT operationalizes this through a balanced conflict strategy where bias signals are equally likely to support correct and incorrect answers, combined with a reward design that penalizes bias-following without rewarding bias agreement.

Researcher Actions

  • Compare this paper against nearby papers in the same arXiv category before using it for protocol decisions.
  • Validate inferred eval signals (Automatic metrics, Simulation environment) against the full paper.
  • Use related-paper links to find stronger protocol-specific references.

Caveats

  • Generated from abstract + metadata only; no PDF parsing.
  • Signals below are heuristic and may miss details reported outside the abstract.

Recommended Queries

Accuracy

Related Papers

Papers are ranked by protocol overlap, extraction signal alignment, and semantic proximity.

No related papers found for this item yet.

Need human evaluators for your AI research? Scale annotation with expert AI Trainers.

Post a Job Get a Quote