Recovering Lesion Parameters from Aphasic Picture Naming Error Profiles in Large Language Models

arXiv cs.CL Papers

Summary

This paper asks whether lesion parameters in LLaVA-Vicuna 13B can be recovered from aphasic picture-naming error profiles. The authors find that perturbation intensity is recoverable while layer index is only approximate, with 81.4% counterfactual fidelity and syndrome-discriminative generalization to stroke survivors, suggesting functional redundancy across transformer layers.

arXiv:2608.06429v1 Announce Type: new Abstract: Interpretability methods for large language models (LLMs) describe internal state but do not directly test whether that state is causally sufficient to produce the observed behavior. In earlier work, we lesioned LLMs to produce error profiles in picture naming, a central task for assessing aphasia, and found that specific lesions produced errors resembling those of individual stroke survivors. Here we ask the inverse question: given an error profile, can the lesion parameters that produced it be recovered, and what does this inverse problem reveal about transformer computation? Lesions in LLaVA-Vicuna 13B were parameterized by layer index, modification percentage, and noise sigma across 4,840 configurations, and error profiles were characterized by a seven-category clinical taxonomy (correct, semantic, unrelated, formal, mixed, neologism, no-response). We trained a multi-task neural network to map error profiles back to perturbation parameters. The problem admitted a partial solution: across 10 independently trained inverse models, modification percentage and noise sigma were recoverable, whereas layer index was recoverable only within a neighborhood. In counterfactual validation, a fresh model instance perturbed with the recovered parameters reproduced the target behavior in 81.4% of cases. This dissociation between low layer recovery and high counterfactual fidelity is consistent with functional redundancy across transformer layers, a property not captured by standard interpretability methods. As an out-of-distribution test, we applied the trained model to picture-naming error profiles from 278 stroke survivors; recovered parameters were syndrome-discriminative, most strongly for perturbation intensity, indicating generalization beyond the training distribution. Counterfactual validation provides a general framework for LLM interpretability claims beyond inverse mapping.
Original Article
View Cached Full Text

Cached at: 08/10/26, 08:01 AM

# Recovering Lesion Parameters from Aphasic Picture Naming Error Profiles in Large Language Models
Source: [https://arxiv.org/abs/2608.06429](https://arxiv.org/abs/2608.06429)
Authors:[Yong Yang](https://arxiv.org/search/cs?searchtype=author&query=Yang,+Y),[Roger Newman\-Norlund](https://arxiv.org/search/cs?searchtype=author&query=Newman-Norlund,+R),[Xiang Guan](https://arxiv.org/search/cs?searchtype=author&query=Guan,+X),[Saeed Ahmadi](https://arxiv.org/search/cs?searchtype=author&query=Ahmadi,+S),[Regan Willis](https://arxiv.org/search/cs?searchtype=author&query=Willis,+R),[Nadra Salman](https://arxiv.org/search/cs?searchtype=author&query=Salman,+N),[Kalil Warren](https://arxiv.org/search/cs?searchtype=author&query=Warren,+K),[Sophie Arheix\-Parras](https://arxiv.org/search/cs?searchtype=author&query=Arheix-Parras,+S),[Srihari Nelakuditi](https://arxiv.org/search/cs?searchtype=author&query=Nelakuditi,+S),[Leonardo Bonilha](https://arxiv.org/search/cs?searchtype=author&query=Bonilha,+L),[Christopher Rorden](https://arxiv.org/search/cs?searchtype=author&query=Rorden,+C),[Rutvik H\. Desai](https://arxiv.org/search/cs?searchtype=author&query=Desai,+R+H),[Julius Fridriksson](https://arxiv.org/search/cs?searchtype=author&query=Fridriksson,+J)

[View PDF](https://arxiv.org/pdf/2608.06429)

> Abstract:Interpretability methods for large language models \(LLMs\) describe internal state but do not directly test whether that state is causally sufficient to produce the observed behavior\. In earlier work, we lesioned LLMs to produce error profiles in picture naming, a central task for assessing aphasia, and found that specific lesions produced errors resembling those of individual stroke survivors\. Here we ask the inverse question: given an error profile, can the lesion parameters that produced it be recovered, and what does this inverse problem reveal about transformer computation? Lesions in LLaVA\-Vicuna 13B were parameterized by layer index, modification percentage, and noise sigma across 4,840 configurations, and error profiles were characterized by a seven\-category clinical taxonomy \(correct, semantic, unrelated, formal, mixed, neologism, no\-response\)\. We trained a multi\-task neural network to map error profiles back to perturbation parameters\. The problem admitted a partial solution: across 10 independently trained inverse models, modification percentage and noise sigma were recoverable, whereas layer index was recoverable only within a neighborhood\. In counterfactual validation, a fresh model instance perturbed with the recovered parameters reproduced the target behavior in 81\.4% of cases\. This dissociation between low layer recovery and high counterfactual fidelity is consistent with functional redundancy across transformer layers, a property not captured by standard interpretability methods\. As an out\-of\-distribution test, we applied the trained model to picture\-naming error profiles from 278 stroke survivors; recovered parameters were syndrome\-discriminative, most strongly for perturbation intensity, indicating generalization beyond the training distribution\. Counterfactual validation provides a general framework for LLM interpretability claims beyond inverse mapping\.

## Submission history

From: Yong Yang \[[view email](https://arxiv.org/show-email/382f819a/2608.06429)\] **\[v1\]**Wed, 5 Aug 2026 20:58:49 UTC \(2,874 KB\)

Similar Articles

Large Vision-Language Models Get Lost in Attention

arXiv cs.AI

This research paper analyzes the internal mechanics of Large Vision-Language Models (LVLMs) using information theory, revealing that attention mechanisms may be redundant while Feed-Forward Networks drive semantic innovation. The authors demonstrate that replacing learned attention weights with random values can yield comparable performance, suggesting current models 'get lost in attention'.

Human-Like Anaphor Resolution in Large Language Models

arXiv cs.CL

This paper investigates whether five open-weight LLMs exhibit human-like sensitivity to psycholinguistic factors in anaphor resolution, using surprisal and comprehension accuracy as behavioral measures. Results show selective cognitive alignment, with some models matching human discourse sensitivity but not semantic interference effects.

Dismantling Pathological Shortcuts: A Causal Framework for Faithful LVLM Decoding

arXiv cs.AI

This paper reveals that hallucination in large vision-language models is caused by a dynamic structural misalignment where certain attention heads act as risky mediators, decoupling from visual evidence to lock onto language priors. The authors propose Fox, a training-free causal intervention framework that diagnoses and physically severs these pathological shortcuts, achieving state-of-the-art performance in faithful decoding.