REVIEW 29 cited by
Function Vectors in Large Language Models
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
Function Vectors in Large Language Models
read the original abstract
We report the presence of a simple neural mechanism that represents an input-output function as a vector within autoregressive transformer language models (LMs). Using causal mediation analysis on a diverse range of in-context-learning (ICL) tasks, we find that a small number attention heads transport a compact representation of the demonstrated task, which we call a function vector (FV). FVs are robust to changes in context, i.e., they trigger execution of the task on inputs such as zero-shot and natural text settings that do not resemble the ICL contexts from which they are collected. We test FVs across a range of tasks, models, and layers and find strong causal effects across settings in middle layers. We investigate the internal structure of FVs and find while that they often contain information that encodes the output space of the function, this information alone is not sufficient to reconstruct an FV. Finally, we test semantic vector composition in FVs, and find that to some extent they can be summed to create vectors that trigger new complex tasks. Our findings show that compact, causal internal vector representations of function abstractions can be explicitly extracted from LLMs. Our code and data are available at https://functions.baulab.info.
Forward citations
Cited by 29 Pith papers
-
The Linear Representation Hypothesis and the Geometry of Large Language Models
Linear representations of high-level concepts in LLMs are formalized via counterfactuals in input and output spaces, unified under a causal inner product that enables consistent probing and steering.
-
Verbalizable Representations Form a Global Workspace in Language Models
Language models represent their current reasoning in a small, readable set of verbalizable vectors (the J-space) that functions like a global workspace.
-
Transformer Field Theory: A Response-Theoretic Approach to Mechanistic Interpretability
Transformer Field Theory frames the residual stream as a field, models patching as source insertion, and uses first-order sensitivities plus Green functions to predict and describe responses, with empirical tests on G...
-
Self-Improving In-Context Learning
A test-time zeroth-order optimization of prompt embeddings using a bounded self-supervised proxy from demonstration log-probabilities improves ICL accuracy and correlates with gains across tasks.
-
Memory Inception: Latent-Space KV Cache Manipulation for Steering LLMs
Memory Inception steers LLMs via selective latent KV cache injection at chosen layers, delivering better control-drift balance than prompting or CAA on personality and reasoning tasks while reducing storage needs.
-
Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior
Manifold steering along activation geometry induces behavioral trajectories matching the natural manifold of outputs, while linear steering produces off-manifold unnatural behaviors.
-
Jailbreaking the Matrix: Nullspace Steering for Controlled Model Subversion
HMNS is a new jailbreak method that uses causal head identification and nullspace-constrained injection to achieve higher attack success rates than prior techniques on aligned language models.
-
Soft Head Selection for Injecting ICL-Derived Task Embeddings
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...
-
Generating Special Triangulations with Transformers
Transformers generate new FRSTs of 4D reflexive polytopes across size ranges and self-improve by retraining on their own outputs.
-
Recoverable but Not Stationary:Local Linear Structures in Weights and Activations
Local low-rank task-gradient structures exist in weights and activations but are non-stationary, with initial recovery updates forming a basis capturing 77% of LoRA displacement and parameter steps aligning 0.58 cosin...
-
Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet
Sparse autoencoders scaled to 34 million features on Claude 3 Sonnet yield interpretable, steerable representations of concrete and abstract concepts that generalize across languages and modalities.
-
Do as I Say, Not as I Do: Instruction-Induction Conflict in LLMs
LLMs show instruction-following rates from 1% to 99% when instructions conflict with hardcoded pattern demonstrations, with output diversity as the main predictor of resistance.
-
Do as I Say, Not as I Do: Instruction-Induction Conflict in LLMs
Experiments reveal that LLMs follow instructions at rates from 1% to 99% when opposed by hardcoded conflicting patterns, with robustness tied to output diversity and alignment with model priors rather than general capability.
-
TFGN: Task-Free, Replay-Free Continual Pre-Training Without Catastrophic Forgetting at LLM Scale
TFGN is an architectural overlay for transformers enabling task-free, replay-free continual pre-training across heterogeneous domains at LLM scale with near-zero backward transfer and high gradient orthogonality.
-
Stories in Space: In-Context Learning Trajectories in Conceptual Belief Space
LLMs perform in-context learning as trajectories through a structured low-dimensional conceptual belief space, with the structure visible in both behavior and internal representations and causally manipulable via inte...
-
Tool Calling is Linearly Readable and Steerable in Language Models
Tool identity is linearly readable and steerable in LLMs via mean activation differences, with 77-100% switch accuracy and error prediction from activation gaps.
-
Memory Inception: Latent-Space KV Cache Manipulation for Steering LLMs
Memory Inception is a training-free method that injects latent KV banks at chosen layers to steer LLMs, achieving superior control-drift balance and up to 118x storage reduction on personality and structured-reasoning tasks.
-
Geometric Deviation as an Unsupervised Pre-Generation Reliability Signal: Probing LLM Representations for Answerability
Geometric deviation of LLM hidden states from an answerable reference centroid provides a pre-generation signal for answerability that works reliably for mathematical prompts (ROC-AUC 0.78-0.84) but not factual ones.
-
The Master Key Hypothesis: Unlocking Cross-Model Capability Transfer via Linear Subspace Alignment
The Master Key Hypothesis states that capabilities are low-dimensional directions transferable across models through linear subspace alignment, with UNLOCK demonstrating gains such as 12.1% accuracy improvement on MAT...
-
How Do Language Models Compose Functions?
LLMs solve compositional factual recall either by computing intermediates or directly, with mechanism choice correlated to translation geometry in embedding spaces.
-
Localizing Task Recognition and Task Learning in In-Context Learning via Attention Head Analysis
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.
-
First-Order Predictable but Pairwise Fragile: Local Task Adaptation in Trained Transformers
Single task-vector perturbations around a multitask LoRA point are first-order linear to 1e-2, but pairwise update-order sensitivity is set by a per-pair Lie bracket and can appear inside that window.
-
Hyper-ICL: Attention Calibration with Hyperbolic Anchor Distillation for Multimodal In-Context Learning
Hyper-ICL learns a query-adaptive low-rank logit adapter plus hyperbolic anchor distillation to reconstruct demonstration effects for demonstration-free multimodal ICL.
-
Decodable but Not Corrected by Fixed Residual-Stream Linear Steering: Evidence from Medical LLM Failure Regimes
Overthinking in medical QA is linearly decodable at 71.6% accuracy yet fixed residual-stream steering yields no correction across 29 configurations, while enabling selective abstention with AUROC 0.610.
-
When Context Sticks: Studying Interference in In-Context Learning
In-context learning shows persistent interference from prior examples, with more misleading linear examples degrading quadratic predictions and training curricula modulating recovery speed.
-
ATLAS: Constitution-Conditioned Latent Geometry and Redistribution Across Language Models and Neural Perturbation Data
ATLAS shows constitutions induce recoverable latent geometry in LLMs that redistributes but remains detectable across models and neural perturbation data via source-defined families and AUC separations.
-
How to use and interpret activation patching
Activation patching provides evidence about neural network circuits when the choice of metric is aligned with the hypothesis and common interpretation errors are avoided.
-
The Hitchhiker's Guide to Agentic AI: From Foundations to Systems
A comprehensive reference book organizing existing techniques for agentic AI systems across LLM substrate, reasoning, agent design patterns, inter-agent coordination, and production deployment.
-
The Hitchhiker's Guide to Agentic AI: From Foundations to Systems
A survey-style reference book mapping the full agentic-AI stack from transformer internals to production deployment, with no new research result.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.