ActMap: Single-Pass Uncertainty Quantification from Generation-Time Activation Maps

• URL: http://arxiv.org/abs/2609.11498v2
• Date: Fri, 11 Sep 2026 17:49:38 GMT
• Title: ActMap: Single-Pass Uncertainty Quantification from Generation-Time Activation Maps
• Abstract summary: ActMap is a representation that compresses the generation-time hidden-state trajectory into a fixed $12times32times128$ of temporal-statistic channels.<n>A lightweight classifier, instantiated as a compact Vision Transformer, reads an estimated correctness probability from each map in a fraction of a millisecond.
• Score: 1.7403133838762448
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Practical uncertainty quantification (UQ) for large language models must decide, from a single generation, whether a specific answer should be trusted. Existing methods either sample multiple generations, read only output-token probabilities, or reduce the model's internal computation to a single hidden state. We introduce ActMap, a white-box representation that compresses the generation-time hidden-state trajectory (every layer, every generated token) into a fixed $12\times32\times128$ tensor of temporal-statistic channels that preserves structure across transformer depth and pooled hidden coordinates. The map is captured during the generation pass with no measurable overhead, has a fixed shape across model depths and hidden sizes, and occupies 96 KiB: a compact artifact that can be retained for audit-relevant generations and probed directly, with occlusion analysis localizing the classifier's signal to mid-depth regions of the map. A lightweight classifier, instantiated as a compact Vision Transformer, reads an estimated correctness probability from each map in a fraction of a millisecond; capacity-matched MLPs perform comparably, indicating the representation itself carries the result. Trained and evaluated in-domain on short-answer QA, direct-answer math, and summarization factuality with three instruction-tuned 7-8B models, ActMap consistently outperforms sampling, token-probability, attention, and embedding baselines, and matches ACT-ViT, a detector trained on dense activation tensors $67\times$ larger, at essentially the same mean AUROC with lower calibration error on ten of twelve pairs. The resulting score supports abstention, routing, and selective verification from a single generation, making it a practical primitive for scalable oversight of deployed models.

Related papers

• Finding Belief Geometries with Sparse Autoencoders [0.0]
  We introduce a pipeline for discovering candidate simplex-structured subspaces in transformer representations.<n>We validate the pipeline on a transformer trained on a hidden Markov model with known belief-state geometry.<n>Preliminary evidence that genuine belief-like geometry exists in Gemma-2-9B's representation space.
  arXiv  Detail & Related papers  (2026-04-03T03:29:48Z)
• CLUE: Non-parametric Verification from Experience via Hidden-State Clustering [64.50919789875233]
  We show that correctness of a solution is encoded as a geometrically separable signature within the trajectory of hidden activations.<n>ClUE consistently outperforms LLM-as-a-judge baselines and matches or exceeds modern confidence-based methods in reranking candidates.
  arXiv  Detail & Related papers  (2025-10-02T02:14:33Z)
• Asynchronously Trained Distributed Topographic Maps [0.0]
  We present an algorithm that uses $N$ autonomous units to generate a feature map by distributed training. Unit autonomy is achieved by sparse interaction in time & space through the combination of a distributed search, and a cascade-driven weight updating scheme.
  arXiv  Detail & Related papers  (2023-01-20T01:15:56Z)
• Localized Randomized Smoothing for Collective Robustness Certification [60.83383487495282]
  We propose a more general collective robustness certificate for all types of models. We show that this approach is beneficial for the larger class of softly local models. The certificate is based on our novel localized randomized smoothing approach.
  arXiv  Detail & Related papers  (2022-10-28T14:10:24Z)
• Self-Supervised Clustering on Image-Subtracted Data with Deep-Embedded Self-Organizing Map [0.0]
  Self-supervised machine learning model, the deep-embedded self-organizing map (DESOM) is applied to real-bogus classification problem. We demonstrate different model training approaches, and find that our best DESOM classifier shows a missed detection rate of 6.6% with a false positive rate of 1.5%.
  arXiv  Detail & Related papers  (2022-09-14T02:37:06Z)
• Constrained Sampling for Class-Agnostic Weakly Supervised Object Localization [10.542859578763068]
  Self-supervised vision transformers can generate accurate localization maps of the objects in an image. We propose leveraging the multiple maps generated by the different transformer heads to acquire pseudo-labels for training a weakly-supervised object localization model.
  arXiv  Detail & Related papers  (2022-09-09T19:58:38Z)
• Discriminative Sampling of Proposals in Self-Supervised Transformers for Weakly Supervised Object Localization [10.542859578763068]
  Self-supervised vision transformers can generate accurate localization maps of the objects in an image. We propose leveraging the multiple maps generated by the different transformer heads to acquire pseudo-labels for training a weakly-supervised object localization model.
  arXiv  Detail & Related papers  (2022-09-09T18:33:23Z)
• Sparse Instance Activation for Real-Time Instance Segmentation [72.23597664935684]
  We propose a conceptually novel, efficient, and fully convolutional framework for real-time instance segmentation. SparseInst has extremely fast inference speed and achieves 40 FPS and 37.9 AP on the COCO benchmark.
  arXiv  Detail & Related papers  (2022-03-24T03:15:39Z)
• CAMERAS: Enhanced Resolution And Sanity preserving Class Activation Mapping for image saliency [61.40511574314069]
  Backpropagation image saliency aims at explaining model predictions by estimating model-centric importance of individual pixels in the input. We propose CAMERAS, a technique to compute high-fidelity backpropagation saliency maps without requiring any external priors.
  arXiv  Detail & Related papers  (2021-06-20T08:20:56Z)
• Graph Sampling Based Deep Metric Learning for Generalizable Person Re-Identification [114.56752624945142]
  We argue that the most popular random sampling method, the well-known PK sampler, is not informative and efficient for deep metric learning. We propose an efficient mini batch sampling method called Graph Sampling (GS) for large-scale metric learning.
  arXiv  Detail & Related papers  (2021-04-04T06:44:15Z)
• Rethinking Localization Map: Towards Accurate Object Perception with Self-Enhancement Maps [78.2581910688094]
  This work introduces a novel self-enhancement method to harvest accurate object localization maps and object boundaries with only category labels as supervision. In particular, the proposed Self-Enhancement Maps achieve the state-of-the-art localization accuracy of 54.88% on ILSVRC.
  arXiv  Detail & Related papers  (2020-06-09T12:35:55Z)
• Generalized Focal Loss: Learning Qualified and Distributed Bounding Boxes for Dense Object Detection [85.53263670166304]
  One-stage detector basically formulates object detection as dense classification and localization. Recent trend for one-stage detectors is to introduce an individual prediction branch to estimate the quality of localization. This paper delves into the representations of the above three fundamental elements: quality estimation, classification and localization.
  arXiv  Detail & Related papers  (2020-06-08T07:24:33Z)

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.