Automated Circuit Interpretation via Probe Prompting

• URL: http://arxiv.org/abs/2511.07002v1
• Date: Mon, 10 Nov 2025 11:53:36 GMT
• Title: Automated Circuit Interpretation via Probe Prompting
• Abstract summary: We present probe prompting, an automated pipeline that transforms attribution graphs into compact, interpretable subgraphs.<n>Across five prompts, probe-prompted subgraphs preserve high explanatory coverage while compressing complexity.<n>Compared to geometric clustering baselines, concept-aligned groups exhibit higher behavioral coherence.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Mechanistic interpretability aims to understand neural networks by identifying which learned features mediate specific behaviors. Attribution graphs reveal these feature pathways, but interpreting them requires extensive manual analysis -- a single prompt can take approximately 2 hours for an experienced circuit tracer. We present probe prompting, an automated pipeline that transforms attribution graphs into compact, interpretable subgraphs built from concept-aligned supernodes. Starting from a seed prompt and target logit, we select high-influence features, generate concept-targeted yet context-varying probes, and group features by cross-prompt activation signatures into Semantic, Relationship, and Say-X categories using transparent decision rules. Across five prompts including classic "capitals" circuits, probe-prompted subgraphs preserve high explanatory coverage while compressing complexity (Completeness 0.83, mean across circuits; Replacement 0.54). Compared to geometric clustering baselines, concept-aligned groups exhibit higher behavioral coherence: 2.3x higher peak-token consistency (0.425 vs 0.183) and 5.8x higher activation-pattern similarity (0.762 vs 0.130), despite lower geometric compactness. Entity-swap tests reveal a layerwise hierarchy: early-layer features transfer robustly (64% transfer rate, mean layer 6.3), while late-layer Say-X features specialize for output promotion (mean layer 16.4), supporting a backbone-and-specialization view of transformer computation. We release code (https://github.com/peppinob-ol/attribution-graph-probing), an interactive demo (https://huggingface.co/spaces/Peppinob/attribution-graph-probing), and minimal artifacts enabling immediate reproduction and community adoption.

Related papers

• How Do Linear Probes Emerge? A Circuit-Tracing Framework with Concept-Targeted Attribution [49.01642339390897]
  Concept-Targeted Attribution (CTA) trains attribution graphs with respect to a linear probe direction.<n>We show that CTA yields probe-specific circuits that explain why an internal concept representation arises in a prompt.
  arXiv  Detail & Related papers  (2026-08-27T08:12:49Z)
• Mapping and Measuring the Behavioral Evolution of Large Language Models [28.603443859513153]
  Benchmark leaderboards summarize how well a language model performs, but not how its behavior relates to that of other models or changes across generations.<n>We characterize the output behavior of 32 models from six families using their responses to a shared bank of 10,000 prompts.
  arXiv  Detail & Related papers  (2026-08-11T15:07:28Z)
• Trajectory Geometry of Transformer Representations Across Layers [0.5599792629509229]
  We recast the transformer forward pass as a discrete population trajectory through a high-dimensional representation manifold.<n>We characterize trajectory geometry using five metrics computed directly in the ambient space.
  arXiv  Detail & Related papers  (2026-06-08T09:54:07Z)
• Beyond Pixels: Introspective and Interactive Grounding for Visualization Agents [3.1626173943755975]
  Vision-Language Models (VLMs) frequently misread values, hallucinate details, and confuse overlapping elements in charts.<n>Current approaches rely solely on pixel interpretation, creating a Pixel-Only Bottleneck.<n>Introspective and Interactive Visual Grounding (IVG) is a framework that combines spec-grounded introspection and view-grounded interaction.
  arXiv  Detail & Related papers  (2026-04-22T22:47:23Z)
• 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)
• Insider Threat Detection Using GCN and Bi-LSTM with Explicit and Implicit Graph Representations [3.294619412118624]
  Insider threat detection (ITD) is challenging due to the subtle and concealed nature of malicious activities performed by trusted users.<n>This paper proposes a post-hoc ITD framework that integrates explicit and implicit graph representations with temporal modelling to capture complex user behaviour patterns.
  arXiv  Detail & Related papers  (2025-12-20T19:48:35Z)
• Cluster Paths: Navigating Interpretability in Neural Networks [3.914630846200639]
  We propose a post-hoc interpretability method that clusters activations at selected layers and represents each input as its sequence of cluster IDs.<n>We show that cluster paths uncover visual concepts, such as color palettes, textures, or object contexts, at multiple network depths.
  arXiv  Detail & Related papers  (2025-10-08T00:41:09Z)
• Federated Spatiotemporal Graph Learning for Passive Attack Detection in Smart Grids [2.721477719641864]
  This paper introduces a graph-centric, multimodal detector that fuses physical-layer and behavioral indicators over temporal windows to detect passive attacks.<n>The model achieves a testing accuracy of 98.32% per-timestep and 93.35% per-sequence at 0.15% FPR.
  arXiv  Detail & Related papers  (2025-09-29T08:52:30Z)
• CLEAR: Conv-Like Linearization Revs Pre-Trained Diffusion Transformers Up [64.38715211969516]
  We introduce a convolution-like local attention strategy termed CLEAR, which limits feature interactions to a local window around each query token.<n>Experiments indicate that, by fine-tuning the attention layer on merely 10K self-generated samples for 10K iterations, we can effectively transfer knowledge from a pre-trained DiT to a student model with linear complexity.
  arXiv  Detail & Related papers  (2024-12-20T17:57:09Z)
• (PASS) Visual Prompt Locates Good Structure Sparsity through a Recurrent HyperNetwork [60.889175951038496]
  Large-scale neural networks have demonstrated remarkable performance in different domains like vision and language processing. One of the key questions of structural pruning is how to estimate the channel significance. We propose a novel algorithmic framework, namely textttPASS. It is a tailored hyper-network to take both visual prompts and network weight statistics as input, and output layer-wise channel sparsity in a recurrent manner.
  arXiv  Detail & Related papers  (2024-07-24T16:47:45Z)
• SegVG: Transferring Object Bounding Box to Segmentation for Visual Grounding [56.079013202051094]
  We present SegVG, a novel method transfers the box-level annotation as signals to provide an additional pixel-level supervision for Visual Grounding. This approach allows us to iteratively exploit the annotation as signals for both box-level regression and pixel-level segmentation.
  arXiv  Detail & Related papers  (2024-07-03T15:30:45Z)
• HyperZ$\cdot$Z$\cdot$W Operator Connects Slow-Fast Networks for Full Context Interaction [0.0]
  Self-attention mechanism utilizes large implicit weight matrices, programmed through dot product-based activations with very few trainable parameters, to enable long sequence modeling. In this paper, we investigate the possibility of discarding residual learning by employing large implicit kernels to achieve full context interaction at each layer of the network. Our model incorporates several innovative components and exhibits excellent properties, such as introducing local feedback error for updating the slow network, stable zero-mean features, faster training convergence, and fewer model parameters.
  arXiv  Detail & Related papers  (2024-01-31T15:57:21Z)
• SimMatchV2: Semi-Supervised Learning with Graph Consistency [53.31681712576555]
  We introduce a new semi-supervised learning algorithm - SimMatchV2. It formulates various consistency regularizations between labeled and unlabeled data from the graph perspective. SimMatchV2 has been validated on multiple semi-supervised learning benchmarks.
  arXiv  Detail & Related papers  (2023-08-13T05:56:36Z)
• Dynamic Perceiver for Efficient Visual Recognition [87.08210214417309]
  We propose Dynamic Perceiver (Dyn-Perceiver) to decouple the feature extraction procedure and the early classification task. A feature branch serves to extract image features, while a classification branch processes a latent code assigned for classification tasks. Early exits are placed exclusively within the classification branch, thus eliminating the need for linear separability in low-level features.
  arXiv  Detail & Related papers  (2023-06-20T03:00:22Z)
• DyG2Vec: Efficient Representation Learning for Dynamic Graphs [26.792732615703372]
  Temporal graph neural networks have shown promising results in learning inductive representations by automatically extracting temporal patterns. We present an efficient yet effective attention-based encoder that leverages temporal edge encodings and window-based subgraph sampling to generate task-agnostic embeddings.
  arXiv  Detail & Related papers  (2022-10-30T18:13:04Z)
• ZippyPoint: Fast Interest Point Detection, Description, and Matching through Mixed Precision Discretization [71.91942002659795]
  We investigate and adapt network quantization techniques to accelerate inference and enable its use on compute limited platforms. ZippyPoint, our efficient quantized network with binary descriptors, improves the network runtime speed, the descriptor matching speed, and the 3D model size. These improvements come at a minor performance degradation as evaluated on the tasks of homography estimation, visual localization, and map-free visual relocalization.
  arXiv  Detail & Related papers  (2022-03-07T18:59:03Z)
• Regularized Densely-connected Pyramid Network for Salient Instance Segmentation [73.17802158095813]
  We propose a new pipeline for end-to-end salient instance segmentation (SIS) To better use the rich feature hierarchies in deep networks, we propose the regularized dense connections. A novel multi-level RoIAlign based decoder is introduced to adaptively aggregate multi-level features for better mask predictions.
  arXiv  Detail & Related papers  (2020-08-28T00:13:30Z)

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.