The paper introduces Sparse Readout Prism (SRP), a method that decomposes a language model’s readout matrix into sparse features, allowing logit‑lens scores to be expressed as sums of feature contributions. SRP reveals that lens readings depend on the corpus used to fit the readout, a phenomenon called corpus conditionality, and that the dominant readout feature remains stable across different corpora. By replacing the original readout with SRP’s sparse approximation, the authors recover 8.9–17.3 percentage points more of the tested logit differences than six geometric‑relation baselines, and ablating features shifts logit differences proportionally to their SRP contributions.
Lens methods interpret large language models (LLMs) by mapping intermediate activations to the output vocabulary, revealing how next-token predictions develop through the network. Trained lenses remain expensive: affine-translator parameters grow quadratically with model width, while exact, full-vocabulary Kullback--Leibler (KL) training dominates memory.
arXiv:2608. 10260v1 Announce Type: new Abstract: Lens methods interpret large language models (LLMs) by mapping intermediate activations to the output vocabulary, revealing how next-token predictions develop through the network.
arXiv:2509. 22415v3 Announce Type: replace-cross Abstract: Multimodal large language models (MLLMs) have achieved strong vision-language performance, yet their token-level visual evidence remains difficult to inspect.
arXiv:2609.00816v1 Announce Type: new Abstract: Scene text recognition is reported as 89--97% accurate on the six standard benchmarks, and the problem is widely treated as saturated. We present an al...
arXiv:2608.31084v1 Announce Type: new Abstract: The Jacobian Lens (J-lens) is a recent tool for interpreting LLMs. It reads a hidden state as a ranked list of vocabulary tokens, leaving multi-token c...
arXiv:2607. 08152v1 Announce Type: cross Abstract: On the recent EyeBench benchmark, predicting reading comprehension from eye movements exposes a stark gap: text-aware models using pretrained language models reach 56--63% AUROC, while gaze-only models operate at chance.
The paper introduces Visual Retrieval Heads (VRHs), a small fraction of attention heads in vision‑language models that are causally responsible for grounding text descriptions to image regions. By recasting head‑scoring methods and evaluating across eleven VLMs and five benchmarks, the authors show that masking the top 20 VRHs can drop grounding accuracy by up to 80 percentage points, while random masking has little effect. VRHs generalize across various visual reference tasks, preserve output format while corrupting localization, and transfer causally across models sharing an LLM backbone.
ReWEIGH the Evidence is a training‑free decoding technique that calibrates token‑level ordinal visual evidence to reduce hallucinations in large vision‑language models. It aggregates vocabulary ranks across visual positions, compares candidates to a token‑specific reference derived from unlabeled images, and applies a bounded penalty only when evidence falls below this reference. Experiments on four 7B backbones show up to a 21.3% reduction in hallucinated object mentions while largely preserving or improving descriptive and general performance, with minimal added latency.
The paper demonstrates that vision‑language models (VLMs) possess a small set of attention heads, called Visual Retrieval Heads (VRHs), that are causally responsible for linking text prompts to specific image regions. By adapting head‑scoring techniques from language models, the authors identify VRHs as the heads whose attention from output prediction tokens, summed over the ground‑truth referent region, most reliably indicates causal grounding. Experiments across eleven VLMs and five referring‑expression benchmarks show that masking the top 20 VRHs can drop grounding accuracy by up to 80 percentage points, while random masking has little effect, and that VRHs generalize across diverse visual tasks and transfer across models sharing an LLM backbone.
arXiv:2608.21819v1 Announce Type: cross Abstract: Reliable image captioning in Vision-Language Models (VLMs) requires captions to be both precise and complete, avoiding unsupported object mentions wh...
arXiv:2607. 23913v1 Announce Type: new Abstract: Modern vision-language models (VLMs) increasingly rely on dynamic or high-resolution visual encoding, producing thousands of visual tokens that substantially increase downstream language-model inference cost.
arXiv:2605. 20950v2 Announce Type: replace-cross Abstract: Vision-Language Models (VLMs) face a bottleneck of prohibitive computational costs arising from massive visual token sequences during inference.