Pith. sign in

REVIEW 10 cited by

In-Context Learning Creates Task Vectors

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 2310.15916 v1 pith:TPL4MHX4 submitted 2023-10-24 cs.CL

classification cs.CL
keywords learningtaskvectorin-contextmodelssingletrainingtransformer
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
abstract

In-context learning (ICL) in Large Language Models (LLMs) has emerged as a powerful new learning paradigm. However, its underlying mechanism is still not well understood. In particular, it is challenging to map it to the "standard" machine learning framework, where one uses a training set $S$ to find a best-fitting function $f(x)$ in some hypothesis class. Here we make progress on this problem by showing that the functions learned by ICL often have a very simple structure: they correspond to the transformer LLM whose only inputs are the query $x$ and a single "task vector" calculated from the training set. Thus, ICL can be seen as compressing $S$ into a single task vector $\boldsymbol{\theta}(S)$ and then using this task vector to modulate the transformer to produce the output. We support the above claim via comprehensive experiments across a range of models and tasks.

Discussion (0). Sign in to comment.

Forward citations

Cited by 10 Pith papers

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

  1. Multimodal Function Vectors for Visual Relations

    cs.AI 2025-10 conditional novelty 6.0 of 10

    Multimodal function vectors extracted from a handful of attention heads in OpenFlamingo-4B encode spatial relations and can be steered, fine-tuned, and composed to improve zero-shot relational reasoning.

  2. Bridging the Knowledge-Prediction Gap in LLMs on Multiple-Choice Questions

    cs.CL 2025-09 conditional novelty 6.0 of 10

    KAPPA reduces the knowledge-prediction gap in LLMs by aligning a prediction-direction coordinate to a knowledge-direction coordinate in the residual stream, yielding accuracy gains on binary-choice MCQs and modest gai...

  3. Train Once, Reuse Everywhere: Generalizable Implicit In-Context Learning by Routing Attention

    cs.CL 2025-09 conditional novelty 6.0 of 10

    ICR extracts shared attention directions from in-context learning and routes them at inference time, enabling zero-shot reuse across tasks.

  4. Enhancing Project-Specific Code Completion by Inferring Internal API Information

    cs.SE 2025-07 conditional novelty 6.0 of 10

    A retrieval-augmented code completion method that infers project-internal APIs from a rough draft and a static knowledge base, beating existing repo-level baselines.

  5. Towards Compute-Optimal Many-Shot In-Context Learning

    cs.CL 2025-07 conditional novelty 6.0 of 10

    Hybrid demonstration selection that adds 20 similar examples to a large cached random or k-means set matches or beats similarity-only selection at up to 10x lower estimated inference cost in many-shot ICL.

  6. Learning without training: The implicit dynamics of in-context learning

    cs.CL 2025-07 conditional novelty 6.0 of 10

    A forward pass with in-context examples is exactly equivalent to a context-free forward pass with a minimal rank-1 update to the MLP weights, with an explicit formula for that update.

  7. Distinct Computations Emerge From Compositional Curricula in In-Context Learning

    cs.LG 2025-06 conditional novelty 6.0 of 10

    When transformer models see easy component examples before a harder combined math problem in one prompt, they solve unseen versions of the combined problem and store intermediate steps internally, unlike models traine...

  8. ICM-Fusion: In-Context Meta-Optimized LoRA Fusion for Multi-Task Adaptation

    cs.CV 2025-08 reject novelty 5.0 of 10

    ICM-Fusion uses a conditional VAE plus task-vector guidance to fuse multiple LoRA adapters into one model, reporting marginal average gains on vision and language benchmarks and larger gains in a few-shot long-tail setup.

  9. Activation Reward Models for Few-Shot Model Alignment

    cs.CV 2025-07 conditional novelty 5.0 of 10

    Mean attention-head activations from a few labeled examples, injected into selected heads, turn a frozen vision-language model into a few-shot reward model that beats prompting and scoring baselines and a new reward-h...

  10. InverseScope: Scalable Activation Inversion for Interpreting Large Language Models

    cs.LG 2025-06 conditional novelty 5.0 of 10

    A new conditional-generation architecture plus a feature-consistency metric make activation inversion practical for LLMs up to 7B parameters, with experiments on IOI, RAVEL, and in-context learning.

Pith tools