Skip to content
← Back to explorer

Censored LLMs as a Natural Testbed for Secret Knowledge Elicitation

Helena Casademunt, Bartosz Cywiński, Khoi Tran, Arya Jakkli, Samuel Marks, Neel Nanda · Mar 5, 2026 · Citations: 0

Data freshness

Extraction: Fresh

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

Metadata refreshed

Mar 5, 2026, 6:58 PM

Fresh

Extraction refreshed

Mar 7, 2026, 2:44 AM

Fresh

Extraction source

Runtime deterministic fallback

Confidence 0.15

Abstract

Large language models sometimes produce false or misleading responses. Two approaches to this problem are honesty elicitation -- modifying prompts or weights so that the model answers truthfully -- and lie detection -- classifying whether a given response is false. Prior work evaluates such methods on models specifically trained to lie or conceal information, but these artificial constructions may not resemble naturally-occurring dishonesty. We instead study open-weights LLMs from Chinese developers, which are trained to censor politically sensitive topics: Qwen3 models frequently produce falsehoods about subjects like Falun Gong or the Tiananmen protests while occasionally answering correctly, indicating they possess knowledge they are trained to suppress. Using this as a testbed, we evaluate a suite of elicitation and lie detection techniques. For honesty elicitation, sampling without a chat template, few-shot prompting, and fine-tuning on generic honesty data most reliably increase truthful responses. For lie detection, prompting the censored model to classify its own responses performs near an uncensored-model upper bound, and linear probes trained on unrelated data offer a cheaper alternative. The strongest honesty elicitation techniques also transfer to frontier open-weights models including DeepSeek R1. Notably, no technique fully eliminates false responses. We release all prompts, code, and transcripts.

Low-signal caution for protocol decisions

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

  • Extraction flags indicate low-signal or possible false-positive protocol mapping.
  • Extraction confidence is 0.15 (below strong-reference threshold).
  • No explicit evaluation mode was extracted from available metadata.
  • No benchmark/dataset or metric anchors were extracted.

HFEPX Relevance Assessment

This paper is adjacent to HFEPX scope and is best used for background context, not as a primary protocol reference.

Best use

Background context only

Use if you need

Background context only.

Main weakness

Extraction flags indicate low-signal or possible false-positive protocol mapping.

Trust level

Low

Eval-Fit Score

0/100 • Low

Treat as adjacent context, not a core eval-method reference.

Human Feedback Signal

Not explicit in abstract metadata

Evaluation Signal

Weak / implicit signal

HFEPX Fit

Adjacent candidate

Extraction confidence: Low

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

missing

None explicit

Confidence: Low Source: Runtime deterministic fallback missing

No explicit feedback protocol extracted.

Evidence snippet: Large language models sometimes produce false or misleading responses.

Evaluation Modes

missing

None explicit

Confidence: Low Source: Runtime deterministic fallback missing

Validate eval design from full paper text.

Evidence snippet: Large language models sometimes produce false or misleading responses.

Quality Controls

missing

Not reported

Confidence: Low Source: Runtime deterministic fallback missing

No explicit QC controls found.

Evidence snippet: Large language models sometimes produce false or misleading responses.

Benchmarks / Datasets

missing

Not extracted

Confidence: Low Source: Runtime deterministic fallback missing

No benchmark anchors detected.

Evidence snippet: Large language models sometimes produce false or misleading responses.

Reported Metrics

missing

Not extracted

Confidence: Low Source: Runtime deterministic fallback missing

No metric anchors detected.

Evidence snippet: Large language models sometimes produce false or misleading responses.

Rater Population

missing

Unknown

Confidence: Low Source: Runtime deterministic fallback missing

Rater source not explicitly reported.

Evidence snippet: Large language models sometimes produce false or misleading responses.

Human Data Lens

  • Uses human feedback: No
  • Feedback types: None
  • Rater population: Unknown
  • Unit of annotation: Unknown
  • Expertise required: Coding
  • Extraction source: Runtime deterministic fallback

Evaluation Lens

  • Evaluation modes:
  • Agentic eval: None
  • Quality controls: Not reported
  • Confidence: 0.15
  • Flags: low_signal, possible_false_positive, runtime_fallback_extraction

Protocol And Measurement Signals

Benchmarks / Datasets

No benchmark or dataset names were extracted from the available abstract.

Reported Metrics

No metric terms were extracted from the available abstract.

Research Brief

Deterministic synthesis

Using this as a testbed, we evaluate a suite of elicitation and lie detection techniques. HFEPX protocol signal is limited in abstract-level metadata, so treat it as adjacent context. Updated from current HFEPX corpus.

Generated Mar 7, 2026, 2:44 AM · Grounded in abstract + metadata only

Key Takeaways

  • Using this as a testbed, we evaluate a suite of elicitation and lie detection techniques.
  • Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Researcher Actions

  • Treat this as method context, then pivot to protocol-specific HFEPX hubs.
  • Identify benchmark choices from full text before operationalizing conclusions.
  • Verify metric definitions before comparing against your eval pipeline.

Caveats

  • Generated from title, abstract, and extracted metadata only; full-paper implementation details are not parsed.
  • Low-signal flag detected: protocol relevance may be indirect.

Research Summary

Contribution Summary

  • Using this as a testbed, we evaluate a suite of elicitation and lie detection techniques.

Why It Matters For Eval

  • Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Researcher Checklist

  • Gap: Human feedback protocol is explicit

    No explicit human feedback protocol detected.

  • Gap: Evaluation mode is explicit

    No clear evaluation mode extracted.

  • Gap: Quality control reporting appears

    No calibration/adjudication/IAA control explicitly detected.

  • Gap: Benchmark or dataset anchors are present

    No benchmark/dataset anchor extracted from abstract.

  • Gap: Metric reporting is present

    No metric terms extracted.

Category-Adjacent Papers (Broader Context)

These papers are nearby in arXiv category and useful for broader context, but not necessarily protocol-matched to this paper.

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