Invariant Gradient Alignment for Robust Reasoning Distillation

• URL: http://arxiv.org/abs/2606.05025v1
• Date: Wed, 03 Jun 2026 15:48:52 GMT
• Title: Invariant Gradient Alignment for Robust Reasoning Distillation
• Abstract summary: Large language models (LLMs) fail on out-of-distribution (OOD) inputs whose semantic surface differs from training data, even when the logical structure is identical.<n>This undermines knowledge distillation pipelines that transfer chain-of-thought reasoning to smaller students.<n>We introduce Invariant Gradient Alignment (IGA), a training framework that aligns gradient updates across semantically diverse but logically isomorphic examples.
• Score: 9.39195684989942
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models (LLMs) suffer from shortcut learning: they systematically fail on out-of-distribution (OOD) inputs whose semantic surface differs from training data, even when the logical structure is identical. This undermines knowledge distillation pipelines that transfer chain-of-thought reasoning to smaller students. We introduce Invariant Gradient Alignment (IGA), a training framework that aligns gradient updates across semantically diverse but logically isomorphic examples via three innovations: (i) Logical Isomer Sets, groups of problems sharing identical logical structure across distinct semantic domains (mathematics, medicine, law, science); (ii) a differentiable \emph{Continuous Gradient Conflict Mask}, that suppresses parameter dimensions with high cross-domain gradient variance while preserving invariant directions; and (iii) a truncated SVD projection of the masked gradient back onto the LoRA low-rank manifold, maintaining parameter efficiency throughout. Theoretically, IGA yields tighter OOD generalization bounds than ERM, scaling with the number of isomer domains, and converges at the standard SGD rate under mild regularity. Empirically, IGA outperforms eight baselines across four benchmarks with accuracy gains up to 14.3 pp over ERM-SFT and a Logical Consistency Score of 0.031 versus 0.142 -- a fourfold improvement in representational invariance.

Related papers

• A*-Thought-V2: Efficient Latent Reasoning via Geometric Dynamics of LLM [40.07171947823021]
  Chain-of-Thought (CoT) improves the reasoning ability of Large Language Models (LLMs) but incurs substantial computation and context costs.<n>We present A*-Thought-V2, a geometric dynamics of LLM guided framework that models CoT as a hidden-state trajectory and replaces hard deletion with an explicit-implicit interleaved latent architecture.
  arXiv  Detail & Related papers  (2026-09-07T17:56:20Z)
• From Concentration to Differentiation and Back: Routing Effective Rank in MoE Reasoning Cohorts [61.695980784491475]
  Test-time scaling produces cohorts of reasoning rollouts, yet there is no standard label-free account of how their internal computation reorganizes as inference unfolds.<n>We introduce routing effective rank deff, the entropy-effective dimensionality of a cross-rollout graph built from MoE expert-routing similarity.
  arXiv  Detail & Related papers  (2026-09-06T05:43:22Z)
• Learning Where Outcomes Change:Credit-Addressable Reasoning for Multimodal Geometry [105.89493240787861]
  We introduce credit-addressable reasoning, in which semantic units exposed during inference also define where learning compares alternatives and assigns credit.<n>We instantiate this principle with Code-CoT, which represents visual relations as line-addressable executable code, and organizes reasoning into typed events.<n>Across nine geometry benchmarks, CE-GRPO achieves an average accuracy of 76.04, outperforming Qwen3-VL-8B and trajectory-level GRPO by $8.09$ and 3.43 points, respectively.
  arXiv  Detail & Related papers  (2026-08-31T08:44:36Z)
• Prototype-Anchored Generalized Manifold Regression for Unknown-Domain Object Detection [58.25418970608328]
  We study Single-Domain Generalized Object Detection (Single-DGOD), which aims to transfer a detector trained on a single source domain to multiple unseen domains.<n>Existing methods mainly rely on simulation-driven strategies, such as data augmentation or textual prompts, to enlarge the training distribution.<n>We propose Manifold Regression with Visual-Text Dual Chain-of-Thought (MR-DCoT), which formulates unknown-domain generalization as a manifold regression problem.
  arXiv  Detail & Related papers  (2026-07-08T09:27:15Z)
• Hierarchical Consistency Learning for Test-time Adaptation in Camouflage Perception [50.278200968044665]
  Camouflaged object detection (COD) aims to localize targets that exhibit minimal perceptual differences from backgrounds through physical attributes.<n>Existing methods, constrained by the static train-then-freeze paradigm, suffer from domain rigidity and dependency annotation.<n>We propose the hierarchical consistency learning framework, which integrates test-time adaptation for dynamic representation recalibration.
  arXiv  Detail & Related papers  (2026-05-25T09:57:46Z)
• SimReg: Achieving Higher Performance in the Pretraining via Embedding Similarity Regularization [55.63565289584336]
  SimReg is an embedding similarity regularization loss that explicitly encourages token representations with the same ground-truth label within each sequence to be more similar.<n>Our analysis reveals that this mechanism introduces gains by enlarging multi-classification margins, thereby enabling more efficient classification.
  arXiv  Detail & Related papers  (2026-05-09T08:59:13Z)
• Preserving Continuous Symmetry in Discrete Spaces: Geometric-Aware Quantization for SO(3)-Equivariant GNNs [12.753341915660073]
  We propose a Geometric-Aware Quantization (GAQ) framework that compresses and accelerates equivariant models.<n>On consumer hardware, GAQ achieves 2.39x inference speedup and 4x memory reduction, enabling stable, energy-conserving molecular dynamics simulations.
  arXiv  Detail & Related papers  (2026-03-05T16:20:21Z)
• Rethinking Diffusion Models with Symmetries through Canonicalization with Applications to Molecular Graph Generation [56.361076943802594]
  CanonFlow achieves state-of-the-art performance on the challenging GEOM-DRUG dataset, and the advantage remains large in few-step generation.
  arXiv  Detail & Related papers  (2026-02-16T18:58:55Z)
• Structured Contrastive Learning for Interpretable Latent Representations [2.8870482999983094]
  We propose Structured Contrastive Learning (SCL), a framework that partitions latent space representations into three semantic groups.<n> Experiments on ECG phase invariance and IMU rotation demonstrate superior performance.<n>This work represents a paradigm shift from reactive data augmentation to proactive structural learning, enabling interpretable latent representations in neural networks.
  arXiv  Detail & Related papers  (2025-11-18T21:18:20Z)
• Understanding Post-Training Structural Changes in Large Language Models [3.054513120350576]
  Post-training fundamentally alters the behavior of large language models (LLMs)<n>This work focuses on two widely adopted post-training methods: instruction tuning and long-chain-of-thought (Long-CoT) distillation.
  arXiv  Detail & Related papers  (2025-09-22T15:03:36Z)
• From Entanglement to Alignment: Representation Space Decomposition for Unsupervised Time Series Domain Adaptation [18.100665738436398]
  We introduce DARSD, a novel UDA framework with theoretical explainability that explicitly realizes UDA tasks from the perspective of representation space decomposition.<n>DarSD consists of three synergistic components: (I) An adversarial learnable common invariant basis that projects original features into a domain-invariant subspace while preserving semantic content; (II) A pseudo-labeling mechanism that dynamically separates target features based on confidence, hindering error accumulation; (III) A hybrid contrastive optimization strategy that simultaneously enforces feature clustering and consistency while mitigating emerging distribution gaps.
  arXiv  Detail & Related papers  (2025-07-28T16:26:28Z)
• Rao-Blackwell Gradient Estimators for Equivariant Denoising Diffusion [55.95767828747407]
  In domains such as molecular and protein generation, physical systems exhibit inherent symmetries that are critical to model.<n>We present a framework that reduces training variance and provides a provably lower-variance gradient estimator.<n>We also present a practical implementation of this estimator incorporating the loss and sampling procedure through a method we call Orbit Diffusion.
  arXiv  Detail & Related papers  (2025-02-14T03:26:57Z)
• PseudoNeg-MAE: Self-Supervised Point Cloud Learning using Conditional Pseudo-Negative Embeddings [55.55445978692678]
  PseudoNeg-MAE enhances global feature representation of point cloud masked autoencoders by making them both discriminative and sensitive to transformations.<n>We propose a novel loss that explicitly penalizes invariant collapse, enabling the network to capture richer transformation cues while preserving discriminative representations.
  arXiv  Detail & Related papers  (2024-09-24T07:57:21Z)

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.