A Circuit for Plural Reference: How LLMs Represent and Retrieve Singular and Plural Entities
cs.CL
Submitted: 2026-09-03
Updated: 2026-09-03
Code: https://github.com/dangthithaoanh/plural-reference-circuit
License: http://creativecommons.org/licenses/by/4.0/
The gist: Coreference resolution is an important task in contextual reasoning.
Terminology
Abstract
Coreference resolution is an important task in contextual reasoning. In this paper, we investigate the mechanism for representing and retrieving singular and plural entities for plural reference. We use a combination of mechanistic interpretability and attention pattern analysis to study the process in which LLMs predict a pronoun to refer back to previously mentioned entities. Using a range of causal intervention techniques, we find a set of attention heads that are responsible for (1) representing coreference information in the input, (2) identifying entities that form a plural reference, (3) transferring the information to the component that is responsible for selecting the antecedents and predicting the pronoun. We also find that LLMs align with humans in preference for plural pronoun. Specifically, entities in a plural construction are more likely to be referred to as a plural entity if they are ontologically similar and are linked by the conjunction "and".
Sources
- On Mechanistic Circuits for Extractive Question-Answering
- Mechanistic Interpretability for AI Safety -- A Review
- A Primer on the Inner Workings of Transformer-based Language Models
- Localizing Model Behavior with Path Patching
- The Llama 3 Herd of Models
- Do Attention Heads in BERT Track Syntactic Dependencies?
- Are Large Language Models Robust Coreference Resolvers?
- Does Circuit Analysis Interpretability Scale? Evidence from Multiple Choice Capabilities in Chinchilla
- The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets
- Fine-Tuning Enhances Existing Mechanisms: A Case Study on Entity Tracking
- Linear Representations of Sentiment in Large Language Models
- Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small
- Qwen3 Technical Report
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering