REVIEW 11 cited by
Attention Heads of Large Language Models: A Survey
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
Attention Heads of Large Language Models: A Survey
read the original abstract
Since the advent of ChatGPT, Large Language Models (LLMs) have excelled in various tasks but remain as black-box systems. Understanding the reasoning bottlenecks of LLMs has become a critical challenge, as these limitations are deeply tied to their internal architecture. Among these, attention heads have emerged as a focal point for investigating the underlying mechanics of LLMs. In this survey, we aim to demystify the internal reasoning processes of LLMs by systematically exploring the roles and mechanisms of attention heads. We first introduce a novel four-stage framework inspired by the human thought process: Knowledge Recalling, In-Context Identification, Latent Reasoning, and Expression Preparation. Using this framework, we comprehensively review existing research to identify and categorize the functions of specific attention heads. Additionally, we analyze the experimental methodologies used to discover these special heads, dividing them into two categories: Modeling-Free and Modeling-Required methods. We further summarize relevant evaluation methods and benchmarks. Finally, we discuss the limitations of current research and propose several potential future directions.
Forward citations
Cited by 11 Pith papers
-
How Do LLMs Read Bug Reports? An Empirical Study of Attention in LLMs for Automated Program Repair
LLM program-repair success is associated with diffuse attention across diagnostic bug-report content, while failures correlate with over-focus on metadata such as version information.
-
CompressKV: Semantic-Retrieval-Guided KV-Cache Compression for Resource-Efficient Long-Context LLM Inference
CompressKV uses Semantic Retrieval Heads to guide KV-cache token selection and layer-wise budget allocation, retaining over 97% performance with 3% cache on LongBench QA tasks.
-
Vocabulary Hijacking in LVLMs: Unveiling Critical Attention Heads by Excluding Inert Tokens to Mitigate Hallucination
LVLMs show vocabulary hijacking by inert tokens that decode to hijacking anchors; HABI locates them, NHAR finds resilient heads, and HAVAE boosts those heads to cut hallucinations.
-
CAST: Mitigating Object Hallucination in Large Vision-Language Models via Caption-Guided Visual Attention Steering
CAST reduces object hallucination in LVLMs by 6.03% on average across five models and five benchmarks by identifying caption-sensitive attention heads and applying optimized steering directions to their outputs, with ...
-
Automated Attention Pattern Discovery at Scale in Large Language Models
AP-MAE reconstructs masked attention patterns in LLMs with high accuracy, generalizes across models, predicts generation correctness at 55-70%, and enables 13.6% accuracy gains via targeted interventions.
-
Steering at the Source: Style Modulation Heads for Robust Persona Control
Steering only three 'Style Modulation Heads' controls personas with less coherency collapse than residual-stream steering, across six personas and two base LLMs.
-
Stop Tracking Me! Proactive Defense Against Attribute Inference Attack in LLMs
TRACE-RPS drops LLM attribute inference accuracy from around 50% to below 5% via fine-grained anonymization plus a two-stage rejection optimization.
-
Localizing Task Recognition and Task Learning in In-Context Learning via Attention Head Analysis
A new framework using Task Subspace Logit Attribution localizes attention heads specialized for task recognition and task learning in in-context learning, showing they align and rotate hidden states within a task subspace.
-
Characterize Then Distill: Mechanistic Reasoning in Large Output Spaces
Reasoning in large output spaces proceeds via shortlisting then fine-grained reasoning; this characterization enables a mechanistic distillation strategy that outperforms standard distillation.
-
CAST: Mitigating Object Hallucination in Large Vision-Language Models via Caption-Guided Visual Attention Steering
A training-free method that steers specific attention heads toward caption-query states reduces object hallucination in LVLMs by ~6% on average across five benchmarks.
-
ART: Attention Replacement Technique to Improve Factuality in LLMs
ART replaces uniform attention in shallow LLM layers with local attention patterns to reduce hallucinations across multiple model architectures.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.