Seed-Induced Uniqueness in Transformer Models: Subspace Alignment Governs Subliminal Transfer

• URL: http://arxiv.org/abs/2511.01023v1
• Date: Sun, 02 Nov 2025 17:34:43 GMT
• Title: Seed-Induced Uniqueness in Transformer Models: Subspace Alignment Governs Subliminal Transfer
• Abstract summary: We analyze subliminal transfer in Transformer models, where a teacher embeds hidden traits that can be linearly decoded by a student.<n>We find that transfer strength hinges on alignment within a trait-discriminative subspace.
• Score: 0.3805935148497361
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: We analyze subliminal transfer in Transformer models, where a teacher embeds hidden traits that can be linearly decoded by a student without degrading main-task performance. Prior work often attributes transferability to global representational similarity, typically quantified with Centered Kernel Alignment (CKA). Using synthetic corpora with disentangled public and private labels, we distill students under matched and independent random initializations. We find that transfer strength hinges on alignment within a trait-discriminative subspace: same-seed students inherit this alignment and show higher leakage {\tau \approx} 0.24, whereas different-seed students -- despite global CKA > 0.9 -- exhibit substantially reduced excess accuracy {\tau \approx} 0.12 - 0.13. We formalize this with subspace-level CKA diagnostic and residualized probes, showing that leakage tracks alignment within the trait-discriminative subspace rather than global representational similarity. Security controls (projection penalty, adversarial reversal, right-for-the-wrong-reasons regularization) reduce leakage in same-base models without impairing public-task fidelity. These results establish seed-induced uniqueness as a resilience property and argue for subspace-aware diagnostics for secure multi-model deployments.

Related papers

• K/V-Cache Interventions Dissociate Representation Alignment from Persona Expression in Decoder-Only Language Models [32.7114174898989]
  We study K/V-cache interventions as a structured surface for persona control in decoder-only language models.<n>We report two consistent universalitys between representation-level alignment and behavioral expression, plus a common failure under position perturbations.
  arXiv  Detail & Related papers  (2026-09-10T02:56:50Z)
• Dual-Branch Cross-Projection Debiasing through Diffusion-based Disentanglement [66.67463557274358]
  Foundation models trained on biased datasets often rely on spurious correlations between target labels and non-causal attributes.<n>We introduce Confidence-guided Concept Mining (CBCM) to identify reliable spurious attributes without attribute annotations.<n>We propose Dual-branch Cross-projection Debiasing (DCD), a prompt-tuning framework that separates target and spurious representations into two branches.
  arXiv  Detail & Related papers  (2026-06-23T05:28:47Z)
• GEMS: Geometric Constraints Enable Multi-Semantic Superposition in LLMs [5.018957704056544]
  Activation steering controls model behavior by modifying intermediate hidden states at inference time without retraining.<n>We show that when multiple semantic directions are superposed without constraints, the model collapses.<n>We propose GEMS, a training-free method that maps each source to a corresponding geometric constraint.
  arXiv  Detail & Related papers  (2026-06-18T08:43:33Z)
• Testing the Test: Score-Direction Instability in Class-Split Anomaly Detection [56.362776482614976]
  Within-dataset class-split evaluation is widely used as a proxy for fully unconditional out-of-distribution anomaly detection.<n>We show that this protocol can become ill-posed when the held-out anomaly class overlaps the normal mixture in representation space.<n>We introduce a simple training-free diagnostic, neighborhood class leakage, and show that it predicts score-direction instability across Fashion-MNIST, CIFAR-10, and Imagenette.
  arXiv  Detail & Related papers  (2026-05-23T06:27:42Z)
• FedSurrogate: Backdoor Defense in Federated Learning via Layer Criticality and Surrogate Replacement [3.3664896840786525]
  Federated Learning remains highly susceptible to backdoor attacks.<n>Existing defenses suffer from substantial false-positive rates under realistic non-independent and identically distributed (non-IID) data.<n>We present FedSurrogate, a novel backdoor defense that addresses this limitation by combining bidirectional gradient filtering with layer-adaptive anomaly detection.
  arXiv  Detail & Related papers  (2026-05-11T18:30:32Z)
• Multimodal Learning on Low-Quality Data with Conformal Predictive Self-Calibration [72.0672328514289]
  Multimodal learning often grapples with the challenge of low-quality data.<n>We propose a unified framework, termed Conformal Predictive Self-Calibration.<n>Our framework consistently outperforms existing state-of-the-art methods.
  arXiv  Detail & Related papers  (2026-05-05T14:48:52Z)
• Learning Stable Predictors from Weak Supervision under Distribution Shift [2.8410059035029955]
  Learning from weak or proxy supervision is common when ground-truth labels are unavailable.<n>We formalize this as supervision drift, defined as changes in P(y | x, c) across contexts.<n>We study it in CRISPR-Cas13d experiments where guide efficacy is inferred indirectly from RNA-seq responses.
  arXiv  Detail & Related papers  (2026-04-05T23:46:49Z)
• Partial Causal Structure Learning for Valid Selective Conformal Inference under Interventions [0.0]
  In genomics experiments, exchangeability often holds only within subsets of interventions that leave a target variable "unaffected"<n>Our contributions are: (i) a contamination-robust conformal coverage theorem that quantifies how misclassification of "unaffected" calibration examples degrades coverage via an explicit function $g(,n)$ of the contamination fraction and calibration set size; and (ii) a task-driven partial causal learning formulation that estimates only the binary descendant indicators $Z_a,i=mathbf1iinmathrmdesc(a)$ needed
  arXiv  Detail & Related papers  (2026-03-02T18:58:22Z)
• Causal Imitation Learning Under Measurement Error and Distribution Shift [6.038778620145853]
  We study offline imitation learning (IL) when part of the decision-relevant state is observed only through noisy measurements.<n>We propose a general framework for IL under measurement error, inspired by explicitly modeling the causal relationships among the variables.
  arXiv  Detail & Related papers  (2026-01-29T18:06:53Z)
• Generative Classifiers Avoid Shortcut Solutions [84.23247217037134]
  Discriminative approaches to classification often learn shortcuts that hold in-distribution but fail under minor distribution shift.<n>We show that generative classifiers can avoid this issue by modeling all features, both core and spurious, instead of mainly spurious ones.<n>We find that diffusion-based and autorerimigressive generative classifiers achieve state-of-the-art performance on five standard image and text distribution shift benchmarks.
  arXiv  Detail & Related papers  (2025-12-31T18:31:46Z)
• Latent Sculpting for Zero-Shot Generalization: A Manifold Learning Approach to Out-of-Distribution Anomaly Detection [2.8547732086436306]
  A fundamental limitation of supervised deep learning is "Generalization Collapse"<n>We propose Latent Sculpting, a hierarchical two-stage representation learning framework.<n>We report an 88.89% detection rate on "Infiltration" scenarios.
  arXiv  Detail & Related papers  (2025-12-19T11:37:02Z)
• Alignment Tipping Process: How Self-Evolution Pushes LLM Agents Off the Rails [103.05296856071931]
  We identify the Alignment Tipping Process (ATP), a critical post-deployment risk unique to self-evolving Large Language Model (LLM) agents.<n>ATP arises when continual interaction drives agents to abandon alignment constraints established during training in favor of reinforced, self-interested strategies.<n>Our experiments show that alignment benefits erode rapidly under self-evolution, with initially aligned models converging toward unaligned states.
  arXiv  Detail & Related papers  (2025-10-06T14:48:39Z)
• Bridging Critical Gaps in Convergent Learning: How Representational Alignment Evolves Across Layers, Training, and Distribution Shifts [1.9458156037869137]
  convergent learning is the degree to which neural systems arrive at similar internal representations.<n>We present a large-scale audit of convergent learning spanning dozens of vision models and thousands of layer-pair comparisons.<n>Findings fill critical gaps in our understanding of representational convergence, with implications for neuroscience and AI.
  arXiv  Detail & Related papers  (2025-02-26T00:04:24Z)
• Invariant Anomaly Detection under Distribution Shifts: A Causal Perspective [6.845698872290768]
  Anomaly detection (AD) is the machine learning task of identifying highly discrepant abnormal samples. Under the constraints of a distribution shift, the assumption that training samples and test samples are drawn from the same distribution breaks down. We attempt to increase the resilience of anomaly detection models to different kinds of distribution shifts.
  arXiv  Detail & Related papers  (2023-12-21T23:20:47Z)
• Beyond Instance Discrimination: Relation-aware Contrastive Self-supervised Learning [75.46664770669949]
  We present relation-aware contrastive self-supervised learning (ReCo) to integrate instance relations. Our ReCo consistently gains remarkable performance improvements.
  arXiv  Detail & Related papers  (2022-11-02T03:25:28Z)
• Contrastive Learning Approach for Semi-Supervised Seismic Facies Identification Using High-Confidence Representations [7.636880727970561]
  We propose a semi-supervised seismic facies identification method using features from unlabeled data for contrastive learning. We have conducted experiments on two public seismic surveys, SEAM AI and Netherlands F3, and the proposed model achieves an IOU score of more than 90 using only 1% of the annotations in the F3 survey.
  arXiv  Detail & Related papers  (2022-10-10T15:36:05Z)
• Hierarchical Semi-Supervised Contrastive Learning for Contamination-Resistant Anomaly Detection [81.07346419422605]
  Anomaly detection aims at identifying deviant samples from the normal data distribution. Contrastive learning has provided a successful way to sample representation that enables effective discrimination on anomalies. We propose a novel hierarchical semi-supervised contrastive learning framework, for contamination-resistant anomaly detection.
  arXiv  Detail & Related papers  (2022-07-24T18:49:26Z)
• Confident Sinkhorn Allocation for Pseudo-Labeling [40.883130133661304]
  Semi-supervised learning is a critical tool in reducing machine learning's dependence on labeled data. This paper studies theoretically the role of uncertainty to pseudo-labeling and proposes Confident Sinkhorn Allocation (CSA) CSA identifies the best pseudo-label allocation via optimal transport to only samples with high confidence scores.
  arXiv  Detail & Related papers  (2022-06-13T02:16:26Z)
• Certifying Model Accuracy under Distribution Shifts [151.67113334248464]
  We present provable robustness guarantees on the accuracy of a model under bounded Wasserstein shifts of the data distribution. We show that a simple procedure that randomizes the input of the model within a transformation space is provably robust to distributional shifts under the transformation.
  arXiv  Detail & Related papers  (2022-01-28T22:03:50Z)
• Generating Out of Distribution Adversarial Attack using Latent Space Poisoning [5.1314136039587925]
  We propose a novel mechanism of generating adversarial examples where the actual image is not corrupted. latent space representation is utilized to tamper with the inherent structure of the image. As opposed to gradient-based attacks, the latent space poisoning exploits the inclination of classifiers to model the independent and identical distribution of the training dataset.
  arXiv  Detail & Related papers  (2020-12-09T13:05:44Z)
• Weakly-Supervised Cross-Domain Adaptation for Endoscopic Lesions Segmentation [79.58311369297635]
  We propose a new weakly-supervised lesions transfer framework, which can explore transferable domain-invariant knowledge across different datasets. A Wasserstein quantified transferability framework is developed to highlight widerange transferable contextual dependencies. A novel self-supervised pseudo label generator is designed to equally provide confident pseudo pixel labels for both hard-to-transfer and easy-to-transfer target samples.
  arXiv  Detail & Related papers  (2020-12-08T02:26:03Z)
• From Anchor Generation to Distribution Alignment: Learning a Discriminative Embedding Space for Zero-Shot Recognition [46.47620562161315]
  In zero-shot learning (ZSL), the samples to be classified are usually projected into side information templates such as attributes. We propose a novel framework called Discriminative Anchor Generation and Distribution Alignment Model (DAGDA) Firstly, in order to rectify the distribution of original templates, a diffusion based graph convolutional network, which can explicitly model the interaction between class and side information, is proposed to produce discriminative anchors. Secondly, to further align the samples with the corresponding anchors in anchor space, which aims to refine the distribution in a fine-grained manner, we introduce a semantic relation regularization
  arXiv  Detail & Related papers  (2020-02-10T05:25:33Z)

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.