MechaRule localizes agonist neurons in LLMs via contrastive hierarchical ablation to ground rule extraction in circuitry, recalling 96.8% of high-effect neurons and reducing task performance when suppressed.
Arithmetic without algo- rithms: Language models solve math with a bag of heuristics.arXiv preprint arXiv:2410.21272
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
years
2026 5verdicts
UNVERDICTED 5roles
background 2polarities
background 2representative citing papers
Transformer represents but does not causally transmit staged algorithmic intermediates for base-digit extraction, diverging from probe predictions.
A prompting method that forces GPAI models to state SE best practices before deciding reduces prompt-induced cognitive biases by 51% on average across eight tested biases.
LLMs show strong spatial generalization to unseen maps in shortest-path tasks but fail length scaling due to recursive instability, with data coverage setting hard limits.
LLMs prioritize surface heuristics such as distance cues over implicit constraints in reasoning tasks, with the new HOB benchmark showing no model exceeds 75% strict accuracy and hints recovering performance.
citing papers explorer
-
Neuron-Anchored Rule Extraction for Large Language Models via Contrastive Hierarchical Ablation
MechaRule localizes agonist neurons in LLMs via contrastive hierarchical ablation to ground rule extraction in circuitry, recalling 96.8% of high-effect neurons and reducing task performance when suppressed.
-
Represented Is Not Computed: A Causal Test of Candidate Algorithmic Intermediates in a Transformer
Transformer represents but does not causally transmit staged algorithmic intermediates for base-digit extraction, diverging from probe predictions.
-
Mitigating Prompt-Induced Cognitive Biases in General-Purpose AI for Software Engineering
A prompting method that forces GPAI models to state SE best practices before deciding reduces prompt-induced cognitive biases by 51% on average across eight tested biases.
-
Generalization in LLM Problem Solving: The Case of the Shortest Path
LLMs show strong spatial generalization to unseen maps in shortest-path tasks but fail length scaling due to recursive instability, with data coverage setting hard limits.
-
The Model Says Walk: How Surface Heuristics Override Implicit Constraints in LLM Reasoning
LLMs prioritize surface heuristics such as distance cues over implicit constraints in reasoning tasks, with the new HOB benchmark showing no model exceeds 75% strict accuracy and hints recovering performance.