Pith. sign in

REVIEW 7 cited by

Do I Know This Entity? Knowledge Awareness and Hallucinations in Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2411.14257 v2 pith:PCSYYJBH submitted 2024-11-21 cs.CL cs.AIcs.LG

Do I Know This Entity? Knowledge Awareness and Hallucinations in Language Models

classification cs.CL cs.AIcs.LG
keywords modelentitydirectionsmodelsautoencoderssparseattributeschat
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Hallucinations in large language models are a widespread problem, yet the mechanisms behind whether models will hallucinate are poorly understood, limiting our ability to solve this problem. Using sparse autoencoders as an interpretability tool, we discover that a key part of these mechanisms is entity recognition, where the model detects if an entity is one it can recall facts about. Sparse autoencoders uncover meaningful directions in the representation space, these detect whether the model recognizes an entity, e.g. detecting it doesn't know about an athlete or a movie. This suggests that models can have self-knowledge: internal representations about their own capabilities. These directions are causally relevant: capable of steering the model to refuse to answer questions about known entities, or to hallucinate attributes of unknown entities when it would otherwise refuse. We demonstrate that despite the sparse autoencoders being trained on the base model, these directions have a causal effect on the chat model's refusal behavior, suggesting that chat finetuning has repurposed this existing mechanism. Furthermore, we provide an initial exploration into the mechanistic role of these directions in the model, finding that they disrupt the attention of downstream heads that typically move entity attributes to the final token.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 7 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. PluRule: A Benchmark for Moderating Pluralistic Communities on Social Media

    cs.CL 2026-05 unverdicted novelty 7.0

    PluRule is a new multimodal multilingual benchmark showing that state-of-the-art vision-language models perform only marginally better than a trivial baseline at detecting specific rule violations in pluralistic onlin...

  2. Diagnosing and Correcting Concept Omission in Multimodal Diffusion Transformers

    cs.CV 2026-05 unverdicted novelty 7.0

    Text embeddings in MM-DiTs contain a detectable omission signal for missing concepts, and amplifying it via OSI reduces concept omission in generated images on FLUX.1-Dev and SD3.5-Medium.

  3. Are LLM Uncertainty and Correctness Encoded by the Same Features? A Functional Dissociation via Sparse Autoencoders

    cs.LG 2026-04 unverdicted novelty 7.0

    Uncertainty and correctness in LLMs are encoded by distinct feature populations, with suppression of confounded features improving accuracy and reducing entropy.

  4. Does Bielik Know What It Doesn't Know? Activation Dispersion Separates Entity Familiarity from Factual Reliability Across Model Scale

    cs.CL 2026-07 conditional novelty 6.0

    Unsupervised MLP activation dispersion separates known from fabricated entities at AUROC 0.95–1.00 across Bielik scales, while factual reliability scales separately and refusals stay near zero.

  5. Diagnosing and Correcting Concept Omission in Multimodal Diffusion Transformers

    cs.CV 2026-05 unverdicted novelty 6.0

    Text embeddings in MM-DiTs encode a detectable omission signal for missing concepts; amplifying it via OSI reduces concept omission in text-to-image outputs on FLUX.1-Dev and SD3.5-Medium.

  6. Why Retrieval-Augmented Generation Fails: A Graph Perspective

    cs.CL 2026-05 unverdicted novelty 6.0

    Attribution graphs reveal that RAG failures arise from shallow fragmented evidence flow in LLMs, enabling topology-based detection and targeted interventions that reinforce question-guided routing.

  7. Per-Entity Bias Mapping for AI Visibility: Why Brand Mentions Require Entity-Specific Calibration

    cs.CL 2026-06 unverdicted novelty 4.0

    Per-Entity Bias Mapping claims aggregate visibility metrics fail because large brands exhibit higher fabricated citation rates than smaller ones in AI responses, attributed to the Brand Hallucination Paradox.