Pith. sign in

REVIEW 7 cited by

What needs to go right for an induction head? A mechanistic study of in-context learning circuits and their formation

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 2404.07129 v1 pith:IVIGOX7U submitted 2024-04-10 cs.LG

What needs to go right for an induction head? A mechanistic study of in-context learning circuits and their formation

classification cs.LG
keywords changephasesubcircuitstrainingformationheadin-contextinduction
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

In-context learning is a powerful emergent ability in transformer models. Prior work in mechanistic interpretability has identified a circuit element that may be critical for in-context learning -- the induction head (IH), which performs a match-and-copy operation. During training of large transformers on natural language data, IHs emerge around the same time as a notable phase change in the loss. Despite the robust evidence for IHs and this interesting coincidence with the phase change, relatively little is known about the diversity and emergence dynamics of IHs. Why is there more than one IH, and how are they dependent on each other? Why do IHs appear all of a sudden, and what are the subcircuits that enable them to emerge? We answer these questions by studying IH emergence dynamics in a controlled setting by training on synthetic data. In doing so, we develop and share a novel optogenetics-inspired causal framework for modifying activations throughout training. Using this framework, we delineate the diverse and additive nature of IHs. By clamping subsets of activations throughout training, we then identify three underlying subcircuits that interact to drive IH formation, yielding the phase change. Furthermore, these subcircuits shed light on data-dependent properties of formation, such as phase change timing, already showing the promise of this more in-depth understanding of subcircuits that need to "go right" for an induction head.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 7 Pith papers

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

  1. Fingerprint, Not Blueprint: How Positional Schemes Set the Default Spectral Algebra of Attention

    cs.LG 2026-07 conditional novelty 7.0

    Positional schemes set the default spectral algebra of attention heads: previous-token heads are rotational under RoPE and content-like under absolute/ALiBi, as a post-function fingerprint rather than a hard constraint.

  2. Why Multimodal In-Context Learning Lags Behind? Unveiling the Inner Mechanisms and Bottlenecks

    cs.CV 2026-04 unverdicted novelty 7.0

    Multimodal ICL lags text-only ICL in few-shot settings due to weak cross-modal reasoning alignment and unreliable task mapping transfer, with an inference-stage method proposed to strengthen transfer.

  3. Soft Head Selection for Injecting ICL-Derived Task Embeddings

    cs.CL 2025-07 conditional novelty 7.0

    SITE applies soft gradient-based head selection to inject ICL-derived task embeddings, outperforming prior embedding adaptation and few-shot ICL across generation, reasoning, and NLU tasks on 12 LLMs from 4B to 70B pa...

  4. Relational reasoning and inductive bias in transformers and large language models

    cs.LG 2025-06 unverdicted novelty 7.0

    In-weights learning induces linear embeddings enabling transitive inference in transformers, whereas in-context learning defaults to match-and-copy unless pre-trained on linear tasks or prompted with linear mental maps.

  5. Belief or Circuitry? Causal Evidence for In-Context Graph Learning

    cs.AI 2026-05 conditional novelty 6.0

    Causal evidence from representation analysis and interventions shows LLMs use both genuine structure inference and induction circuits in parallel for in-context graph learning.

  6. Localizing Task Recognition and Task Learning in In-Context Learning via Attention Head Analysis

    cs.CL 2025-09 unverdicted novelty 6.0

    A new framework using Task Subspace Logit Attribution localizes attention heads specialized for task recognition and task learning in in-context learning, showing they align and rotate hidden states within a task subspace.

  7. Mechanistic Interpretability for Neural Networks: Circuits, Sparse Features and Symbolic Reasoning

    cs.LG 2026-07 accept

    A scoping review surveying circuit analysis, sparse autoencoders, activation steering, and neurosymbolic frameworks for interpreting and controlling Transformer-based neural networks.