AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Hedonic Neurons: A Mechanistic Mapping of Latent Coalitions in Transformer MLPs

Tanya Chowdhury, Atharva Nijasure, Yair Zick, James Allan

ICLR 2026 · Architectures and efficiency › RNNs and normalization · approach: Layer-wise analysis of language models

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use PAS (ablation-based) valuations instead of OCA in hedonic coalition discovery (ablation)Hedonic coalitions (LLaMA-3.1) · CQTR OOD dropOOD drop: 0.07 -> 0.11Improved
use hedonic coalitions instead of K-means or hierarchical clustering (ablation)neuron coalition builder · OOD drop on CQTROOD drop: 0.02-0.03 -> 0.11-0.14Improved
use hedonic coalitions instead of clustering for feature alignment (ablation)neuron coalition builder · MSLR IR heuristic alignmentalignment R^2: 0.11-0.18 -> 0.55-0.67Improved
use hedonic coalitions as macro-features for ridge regression (ablation)coalition predictivity · DL-19/20 OOD, avg of 3 LLMsridge R^2: 0.15-0.21 -> 0.43/0.42/0.47Improved
use hedonic coalitions instead of spherical K-means for synergy (ablation)neuron coalition builder · pairwise and ratio synergy, 3 LLMs x 3 taskssynergy: +0.29 Pairwise, +0.49 RatioImproved
use random neuron subsets rather than clusters (ablation)neuron coalition builder · pairwise synergypairwise synergy: 0.01 -> -0.23 for K-means on CQTR LLaMA vs RandomHurt
track coalitions across consecutive layers 7 to 14cross-layer coalition tracking · CQTR/Mean-TF/L/RM, 3 LLMscoalition fate percentage: vanish 60-75%, persist <12%, merges near 0%Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy