Layer Specialization Underlying Compositional Reasoning in Transformers

• URL: http://arxiv.org/abs/2510.17469v1
• Date: Mon, 20 Oct 2025 12:08:22 GMT
• Title: Layer Specialization Underlying Compositional Reasoning in Transformers
• Authors: Jing Liu,
• Abstract summary: Transformers exhibit compositional reasoning on sequences not observed during training.<n>We investigate this phenomenon using the Random Hierarchy Model (RHM), a probabilistic context-free grammar.<n>Mechanistically, we identify a progressive emergence of layer specialization during training that correlates with generalization performance.
• Score: 6.581088182267414
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Transformers exhibit compositional reasoning on sequences not observed during training, a capability often attributed to in-context learning (ICL) and skill composition. We investigate this phenomenon using the Random Hierarchy Model (RHM), a probabilistic context-free grammar that generates sequences through recursive rule application. Models are trained on subsets of sequences and evaluated across four generalization conditions: memorization, in-distribution generalization, out-of-distribution generalization with the same rules, and cross-layer transfer. Behaviorally, performance improves systematically with task complexity and the number of in-context examples, with out-of-distribution tasks requiring substantially more examples than in-distribution scenarios. Mechanistically, we identify a progressive emergence of layer specialization during training that correlates with generalization performance. Principal component analysis and attention pattern clustering reveal that transformers develop structured, hierarchically organized representations in specialized layers. These results demonstrate that transformers develop modular, interpretable mechanisms supporting compositional reasoning, linking internal algorithmic structure to observed behavioral capabilities.

Related papers

• Unlocking Out-of-Distribution Generalization in Transformers via Recursive Latent Space Reasoning [50.99796659680724]
  This work investigates out-of-distribution (OOD) generalization in Transformer networks using a GSM8K-style modular arithmetic on computational graphs task as a testbed.<n>We introduce and explore a set of four architectural mechanisms aimed at enhancing OOD generalization.<n>We complement these empirical results with a detailed mechanistic interpretability analysis that reveals how these mechanisms give rise to robust OOD generalization abilities.
  arXiv  Detail & Related papers  (2025-10-15T21:03:59Z)
• Propositional Logic for Probing Generalization in Neural Networks [3.6037930269014633]
  We investigate the generalization behavior of three key neural architectures (Transformers, Graph Convolution Networks and LSTMs) in a controlled task rooted in propositional logic.<n>We find thatTransformers fail to apply negation compositionally, unless structural biases are introduced.<n>Our findings highlight persistent limitations in the ability of standard architectures to learn systematic representations of logical operators.
  arXiv  Detail & Related papers  (2025-06-10T16:46:05Z)
• How do Transformers Learn Implicit Reasoning? [67.02072851088637]
  We study how implicit multi-hop reasoning emerges by training transformers from scratch in a controlled symbolic environment.<n>We find that training with atomic triples is not necessary but accelerates learning, and that second-hop generalization relies on query-level exposure to specific compositional structures.
  arXiv  Detail & Related papers  (2025-05-29T17:02:49Z)
• Dynamics of Transient Structure in In-Context Linear Regression Transformers [0.5242869847419834]
  We show that when transformers are trained on in-context linear regression tasks with intermediate task diversity, they behave like ridge regression before specializing to the tasks in their training distribution.<n>This transition from a general solution to a specialized solution is revealed by joint trajectory principal component analysis.<n>We empirically validate this explanation by measuring the model complexity of our transformers as defined by the local learning coefficient.
  arXiv  Detail & Related papers  (2025-01-29T16:32:14Z)
• Interpreting Affine Recurrence Learning in GPT-style Transformers [54.01174470722201]
  In-context learning allows GPT-style transformers to generalize during inference without modifying their weights. This paper focuses specifically on their ability to learn and predict affine recurrences as an ICL task. We analyze the model's internal operations using both empirical and theoretical approaches.
  arXiv  Detail & Related papers  (2024-10-22T21:30:01Z)
• Principled Understanding of Generalization for Generative Transformer Models in Arithmetic Reasoning Tasks [5.522116934552708]
  Transformer-based models excel in various tasks but their generalization capabilities, especially in arithmetic reasoning, remain incompletely understood.<n>This paper develops a unified theoretical framework for understanding the generalization behaviors of transformers in arithmetic tasks.
  arXiv  Detail & Related papers  (2024-07-25T11:35:22Z)
• Strengthening Structural Inductive Biases by Pre-training to Perform Syntactic Transformations [75.14793516745374]
  We propose to strengthen the structural inductive bias of a Transformer by intermediate pre-training. Our experiments confirm that this helps with few-shot learning of syntactic tasks such as chunking. Our analysis shows that the intermediate pre-training leads to attention heads that keep track of which syntactic transformation needs to be applied to which token.
  arXiv  Detail & Related papers  (2024-07-05T14:29:44Z)
• Learning Syntax Without Planting Trees: Understanding Hierarchical Generalization in Transformers [74.96551626420188]
  Transformers trained on natural language data have been shown to learn its hierarchical structure and generalize to sentences with unseen syntactic structures.<n>We investigate sources of inductive bias in transformer models and their training that could cause such generalization behavior to emerge.
  arXiv  Detail & Related papers  (2024-04-25T07:10:29Z)
• Inducing Systematicity in Transformers by Attending to Structurally Quantized Embeddings [60.698130703909804]
  Transformers generalize to novel compositions of structures and entities after being trained on a complex dataset. We propose SQ-Transformer that explicitly encourages systematicity in the embeddings and attention layers. We show that SQ-Transformer achieves stronger compositional generalization than the vanilla Transformer on multiple low-complexity semantic parsing and machine translation datasets.
  arXiv  Detail & Related papers  (2024-02-09T15:53:15Z)
• Systematic Generalization and Emergent Structures in Transformers Trained on Structured Tasks [6.525090891505941]
  We show how a causal transformer can perform a set of algorithmic tasks, including copying, sorting, and hierarchical compositions. We show that two-layer transformers learn generalizable solutions to multi-level problems and develop signs of systematic task decomposition. These results provide key insights into how transformer models may be capable of decomposing complex decisions into reusable, multi-level policies.
  arXiv  Detail & Related papers  (2022-10-02T00:46:36Z)

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.