Universal Response and Emergence of Induction in LLMs

• URL: http://arxiv.org/abs/2411.07071v1
• Date: Mon, 11 Nov 2024 15:47:15 GMT
• Title: Universal Response and Emergence of Induction in LLMs
• Authors: Niclas Luick,
• Abstract summary: We study the emergence of induction behavior within LLMs by probing their response to weak single-token perturbations of the residual stream. We find that LLMs exhibit a robust, universal regime in which their response remains scale-invariant under changes in perturbation strength. Our results provide insights into the collective interplay of components within LLMs and serve as a benchmark for large-scale circuit analysis.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: While induction is considered a key mechanism for in-context learning in LLMs, understanding its precise circuit decomposition beyond toy models remains elusive. Here, we study the emergence of induction behavior within LLMs by probing their response to weak single-token perturbations of the residual stream. We find that LLMs exhibit a robust, universal regime in which their response remains scale-invariant under changes in perturbation strength, thereby allowing us to quantify the build-up of token correlations throughout the model. By applying our method, we observe signatures of induction behavior within the residual stream of Gemma-2-2B, Llama-3.2-3B, and GPT-2-XL. Across all models, we find that these induction signatures gradually emerge within intermediate layers and identify the relevant model sections composing this behavior. Our results provide insights into the collective interplay of components within LLMs and serve as a benchmark for large-scale circuit analysis.

Related papers

• Out-of-Distribution Generalization of In-Context Learning: A Low-Dimensional Subspace Perspective [9.249642973141107]
  We demystify the out-of-distribution capabilities of in-context learning (ICL) by studying linear regression tasks parameterized with low-rank covariance matrices.<n>We prove that a single-layer linear attention model incurs a test risk with a non-negligible dependence on the angle, illustrating that ICL is not robust to such distribution shifts.<n>This suggests that the OOD generalization ability of Transformers may actually stem from the new task lying within the span of those encountered during training.
  arXiv  Detail & Related papers  (2025-05-20T18:15:49Z)
• Analyzing the Role of Permutation Invariance in Linear Mode Connectivity [4.516746821973374]
  We show that as the student network width $m$ increases, the LMC loss barrier modulo permutation exhibits a double descent behavior. We observe a sharp transition in the sparsity of GD/SGD solutions when increasing the learning rate and investigate how this sparsity preference affects the LMC loss barrier modulo permutation.
  arXiv  Detail & Related papers  (2025-03-08T01:12:27Z)
• Satori: Reinforcement Learning with Chain-of-Action-Thought Enhances LLM Reasoning via Autoregressive Search [57.28671084993782]
  Large language models (LLMs) have demonstrated remarkable reasoning capabilities across diverse domains. Recent studies have shown that increasing test-time computation enhances LLMs' reasoning capabilities. We propose a two-stage training paradigm: 1) a small-scale format tuning stage to internalize the COAT reasoning format and 2) a large-scale self-improvement stage leveraging reinforcement learning.
  arXiv  Detail & Related papers  (2025-02-04T17:26:58Z)
• Mitigating Forgetting in LLM Fine-Tuning via Low-Perplexity Token Learning [61.99353167168545]
  We show that fine-tuning with LLM-generated data improves target task performance and reduces non-target task degradation.<n>This is the first work to provide an empirical explanation based on token perplexity reduction to mitigate catastrophic forgetting in LLMs after fine-tuning.
  arXiv  Detail & Related papers  (2025-01-24T08:18:56Z)
• Reinforcement Learning for Aligning Large Language Models Agents with Interactive Environments: Quantifying and Mitigating Prompt Overfitting [40.78026627009521]
  Reinforcement learning (RL) is a promising approach for aligning large language models (LLMs) knowledge with sequential decision-making tasks. We propose a novel framework to analyze the sensitivity of LLMs to prompt formulations following RL training in a textual environment.
  arXiv  Detail & Related papers  (2024-10-25T18:25:35Z)
• Understanding Ranking LLMs: A Mechanistic Analysis for Information Retrieval [20.353393773305672]
  We employ a probing-based analysis to examine neuron activations in ranking LLMs. Our study spans a broad range of feature categories, including lexical signals, document structure, query-document interactions, and complex semantic representations. Our findings offer crucial insights for developing more transparent and reliable retrieval systems.
  arXiv  Detail & Related papers  (2024-10-24T08:20:10Z)
• Mamba-PTQ: Outlier Channels in Recurrent Large Language Models [49.1574468325115]
  We show that Mamba models exhibit the same pattern of outlier channels observed in attention-based LLMs. We show that the reason for the difficulty of quantizing SSMs is caused by activation outliers, similar to those observed in transformer-based LLMs.
  arXiv  Detail & Related papers  (2024-07-17T08:21:06Z)
• Induction Heads as an Essential Mechanism for Pattern Matching in In-context Learning [12.911829891263263]
  We show that even a minimal ablation of induction heads leads to ICL performance decreases of up to 32% for abstract pattern recognition tasks. For NLP tasks, this ablation substantially decreases the model's ability to benefit from examples, bringing few-shot ICL performance close to that of zero-shot prompts.
  arXiv  Detail & Related papers  (2024-07-09T16:29:21Z)
• Quantifying Emergence in Large Language Models [31.608080868988825]
  We propose a quantifiable solution for estimating emergence of LLMs. Inspired by emergentism in dynamics, we quantify the strength of emergence by comparing the entropy reduction of the macroscopic (semantic) level with that of the microscopic (token) level. Our method demonstrates consistent behaviors across a suite of LMs under both in-context learning (ICL) and natural sentences.
  arXiv  Detail & Related papers  (2024-05-21T09:12:20Z)
• Sparsity-Guided Holistic Explanation for LLMs with Interpretable Inference-Time Intervention [53.896974148579346]
  Large Language Models (LLMs) have achieved unprecedented breakthroughs in various natural language processing domains. The enigmatic black-box'' nature of LLMs remains a significant challenge for interpretability, hampering transparent and accountable applications. We propose a novel methodology anchored in sparsity-guided techniques, aiming to provide a holistic interpretation of LLMs.
  arXiv  Detail & Related papers  (2023-12-22T19:55:58Z)
• On the Relation between Internal Language Model and Sequence Discriminative Training for Neural Transducers [52.88268942796418]
  Internal language model (ILM) subtraction has been widely applied to improve the performance of the RNN-Transducer. We show that sequence discriminative training has a strong correlation with ILM subtraction from both theoretical and empirical points of view.
  arXiv  Detail & Related papers  (2023-09-25T13:35:28Z)
• Layer-wise Feedback Propagation [53.00944147633484]
  We present Layer-wise Feedback Propagation (LFP), a novel training approach for neural-network-like predictors. LFP assigns rewards to individual connections based on their respective contributions to solving a given task. We demonstrate its effectiveness in achieving comparable performance to gradient descent on various models and datasets.
  arXiv  Detail & Related papers  (2023-08-23T10:48:28Z)
• An Empirical Study of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning [70.48605869773814]
  Catastrophic forgetting (CF) is a phenomenon that occurs in machine learning when a model forgets previously learned information. This study empirically evaluates the forgetting phenomenon in large language models during continual instruction tuning.
  arXiv  Detail & Related papers  (2023-08-17T02:53:23Z)
• Counterfactual Maximum Likelihood Estimation for Training Deep Networks [83.44219640437657]
  Deep learning models are prone to learning spurious correlations that should not be learned as predictive clues. We propose a causality-based training framework to reduce the spurious correlations caused by observable confounders. We conduct experiments on two real-world tasks: Natural Language Inference (NLI) and Image Captioning.
  arXiv  Detail & Related papers  (2021-06-07T17:47:16Z)
• Extreme Memorization via Scale of Initialization [72.78162454173803]
  We construct an experimental setup in which changing the scale of initialization strongly impacts the implicit regularization induced by SGD. We find that the extent and manner in which generalization ability is affected depends on the activation and loss function used. In the case of the homogeneous ReLU activation, we show that this behavior can be attributed to the loss function.
  arXiv  Detail & Related papers  (2020-08-31T04:53:11Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.