Pith. sign in

REVIEW 6 cited by

Interpretability in Parameter Space: Minimizing Mechanistic Description Length with Attribution-based Parameter Decomposition

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

arxiv 2501.14926 v4 pith:HZR6AAVJ submitted 2025-01-24 cs.LG stat.ML

classification cs.LGstat.ML
keywords networkmechanisticneuralcomponentsdecompositionidentifyinginterpretabilitymechanisms
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Mechanistic interpretability aims to understand the internal mechanisms learned by neural networks. Despite recent progress toward this goal, it remains unclear how best to decompose neural network parameters into mechanistic components. We introduce Attribution-based Parameter Decomposition (APD), a method that directly decomposes a neural network's parameters into components that (i) are faithful to the parameters of the original network, (ii) require a minimal number of components to process any input, and (iii) are maximally simple. Our approach thus optimizes for a minimal length description of the network's mechanisms. We demonstrate APD's effectiveness by successfully identifying ground truth mechanisms in multiple toy experimental settings: Recovering features from superposition; separating compressed computations; and identifying cross-layer distributed representations. While challenges remain to scaling APD to non-toy models, our results suggest solutions to several open problems in mechanistic interpretability, including identifying minimal circuits in superposition, offering a conceptual foundation for 'features', and providing an architecture-agnostic framework for neural network decomposition.

Discussion (0). Sign in to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Compressed Computation under $L^4$ Loss is likely Computation in Superposition

    cs.LG 2026-07 accept novelty 6.5 of 10

    Training a 50-neuron ReLU network under L4 loss elicits sparse binary codewords over neurons that compute 100 sparse ReLUs in superposition, recovered by a three-scalar ansatz.

  2. Individual Parameters in Weight-Sparse Transformers Appear Interpretable

    cs.LG 2026-07 conditional novelty 6.5 of 10

    An automated LLM pipeline finds that 12–31% of nonzero weights in weight-sparse transformers admit short, held-out-validated descriptions of when they matter, far above dense controls.

  3. Targeted Recovery of Weight-Space Mechanisms From Neural Networks

    cs.LG 2026-06 conditional novelty 6.0 of 10

    A targeted decomposition method recovers the weight-space mechanisms behind specific inputs at low FLOPs, enabling focused ablation and rewiring of a 12-block transformer.

  4. Compressed Computation: Dense Circuits in a Toy Model of the Universal-AND Problem

    cs.LG 2025-07 conditional novelty 6.0 of 10

    A toy model learns a dense binary-weighted circuit for the Universal-AND problem, which outperforms prior sparse constructions at low sparsity.

  5. Stochastic Parameter Decomposition

    cs.LG 2025-06 conditional novelty 6.0 of 10

    SPD uses stochastic masking and a learned causal importance function to decompose neural network parameters into sparsely active rank-one subcomponents, recovering ground-truth mechanisms in toy models where APD struggled.

  6. Distribution-Aware Feature Selection for SAEs

    cs.LG 2025-08 conditional novelty 5.0 of 10

    Sampled-SAE pre-selects a candidate pool of features using batch-level norms or entropy before batch top-K, creating a tunable family that trades reconstruction fidelity for improved probing and reduced absorption on ...

Pith tools