Pith. sign in

REVIEW 4 cited by

Do LLMs dream of elephants (when told not to)? Latent concept association and associative memory in transformers

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 2406.18400 v2 pith:75HV63KM submitted 2024-06-26 cs.CL cs.LGstat.ML

classification cs.CLcs.LGstat.ML
keywords llmsmemoryassociativefactsassociationconceptcontextslatent
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

Large Language Models (LLMs) have the capacity to store and recall facts. Through experimentation with open-source models, we observe that this ability to retrieve facts can be easily manipulated by changing contexts, even without altering their factual meanings. These findings highlight that LLMs might behave like an associative memory model where certain tokens in the contexts serve as clues to retrieving facts. We mathematically explore this property by studying how transformers, the building blocks of LLMs, can complete such memory tasks. We study a simple latent concept association problem with a one-layer transformer and we show theoretically and empirically that the transformer gathers information using self-attention and uses the value matrix for associative memory.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 4 Pith papers

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

  1. Stochastic Chameleons: Irrelevant Context Hallucinations Reveal Class-Based (Mis)Generalization in LLMs

    cs.CL 2025-05 conditional novelty 6.0 of 10

    LLMs systematically combine abstract category cues from a query with features from irrelevant context, causing structured answer flips, a behavior the authors call class-based (mis)generalization.

  2. Associative memory inspires improvements for in-context learning using a novel attention residual stream architecture

    cs.NE 2024-12 conditional novelty 6.0 of 10

    A residual connection that copies previous-layer attention values into the current layer improves in-context learning in transformers up to 1B parameters.

  3. Rethinking Associative Memory Mechanism in Induction Head

    cs.CL 2024-12 conditional novelty 6.0 of 10

    A two-layer transformer with relative positional encoding keeps its induction head active across the whole sequence, while absolute positional encoding loses it in the second half.

  4. Understanding Factual Recall in Transformers via Associative Memories

    cs.LG 2024-12 accept novelty 6.0 of 10

    A one-layer transformer can store facts at near-optimal capacity by using attention value matrices or an MLP as associative memories, and training passes through a hallucination stage.

Pith tools