interpretability

topic21 events

papersTODAY 04:00 UTC

arXiv Paper Proposes Method for Diversified Counterfactual Explanations

A new arXiv preprint describes an approach for generating counterfactual examples that are both varied and human-interpretable, drawing on expert knowledge to guide the search. Counterfactual examples are a common technique in explainable AI, since they show the smallest input changes that would flip a model's prediction. The work aims to address the limited diversity typical of existing methods while keeping the resulting explanations understandable.

explainability/xai/counterfactuals/research

papersTODAY 04:00 UTC

Study locates and steers opportunity-recognition behavior inside LLMs

A new arXiv paper examines how entrepreneurial cognition research can be extended to large language models, which are increasingly used in entrepreneurial tasks. The authors identify an internal representation tied to opportunity recognition and show they can causally steer it, effectively turning the behavior up or down. The work sits at the intersection of entrepreneurship theory and interpretability research on model internals.

research/interpretability/llm/entrepreneurship

papersTODAY 04:00 UTC

arXiv paper proposes scalable neuro-symbolic approach to generative interpretability

A new arXiv preprint argues that post-hoc interpretability is inadequate as language models shift from chatbots to agentic systems whose actions can have irreversible real-world effects. The authors propose a neuro-symbolic framework aimed at making model reasoning interpretable by construction rather than explained after the fact, and claim it scales. The work is a research preprint and has not been peer reviewed.

interpretability/neuro-symbolic/agents/arxiv

papersTODAY 04:00 UTC

Audio language models track speakers via text backbone attention, study finds

A new study examines how audio language models attribute speech to the correct speaker, finding accuracy of only 6 to 16 percent on a six-speaker task, below random guessing. The authors show that speaker tracking relies on attention heads in the model's text backbone, and that altering a subset of those heads shifts which speaker the model retrieves.

audio/attention/multimodal/research

papersTODAY 04:00 UTC

Forked Futures Method Tests Reusable Causal Interfaces in Language Models

A new arXiv paper argues that probing a language model's current answer is not enough to show it has a stable, reusable internal interface, since the same output can come from hidden states that would support different later computations. The authors propose a "forked futures" approach, in which future operations are sampled only after the fact, to test whether internal representations serve as causal interfaces that transfer across tasks. The work targets interpretability and evaluation of model internals rather than a product release.

interpretability/research/llm-internals/evaluation

papersTODAY 04:00 UTC

Paper Proposes EventGraph and EventField Pipeline for Interpretable Temporal Video Reasoning

A new arXiv preprint describes a video reasoning approach that pairs a discrete event graph with a continuous event field, plus a human-readable glyph view, so intermediate reasoning steps can be inspected. The authors evaluate the pipeline on a curated EPIC-KITCHENS subset containing 10 videos and 50 questions about temporal relationships. The work sits in the interpretability and video-language research space rather than announcing a product or model release.

video-reasoning/interpretability/multimodal/benchmarks

papersTODAY 04:00 UTC

Paper explores hierarchical structure of speaker embeddings in recognition networks

A new arXiv preprint examines how speaker recognition networks organise the latent representations they learn from speech utterances. The work targets the limited transparency of these models by analysing whether speaker embeddings are arranged hierarchically. The authors position the study as a step toward explaining the internal mechanisms behind speaker identification systems.

speech/interpretability/embeddings/research

papersTODAY 04:00 UTC

EEG-Xplain framework targets interpretability of EEG foundation models

A new arXiv paper proposes EEG-Xplain, a unified attribution framework intended to make EEG foundation models such as BIOT, LaBraM, and EEGMamba more interpretable. The authors argue that the black-box nature of these models hinders clinical trust and neuroscientific validation. The work aims to provide a common approach for attributing model outputs to neural signal inputs.

interpretability/eeg/foundation-models/research

papersTODAY 04:00 UTC

arXiv paper proposes traceable multi-hop navigation for knowledge graph question answering

A new arXiv preprint introduces an approach for multi-hop knowledge graph question answering that emphasizes how a model travels through a graph rather than only the answer it produces. The authors argue that prior systems typically optimize for final-answer accuracy, leaving the relational evidence path unexplained. The work, titled "Theseus in the Graph," aims to make those navigation steps traceable.

knowledge-graphs/question-answering/interpretability/research

papersTODAY 04:00 UTC

Paper Proposes Using Model Internals to Predict Behavior on Unseen Data

A new arXiv paper reframes interpretability research around predicting how a model will respond to previously unseen inputs, rather than only to targeted mechanistic interventions. The authors use a model's internal representations to forecast its out-of-distribution behavior. The work appears in two arXiv listings, cs.AI and cs.LG, as a replacement submission.

interpretability/research/arxiv/evaluation

papersTODAY 04:00 UTC

Paper reviews tensorization for neural network compression and interpretability

A revised arXiv paper examines tensorization, a method that reshapes a network's dense weight matrices into higher-order tensors and approximates them with low-rank tensor network decompositions. The authors argue the approach remains underused despite promising results as a model compression technique, and they highlight its potential for making networks easier to interpret. The submission appears as a replacement cross-list across arXiv's AI and machine learning categories.

research/model-compression/interpretability/arxiv

papersTODAY 04:00 UTC

Bypass Observation: Read-Only Layer-Wise Semantic Extraction for LLMs

A new arXiv paper proposes Bypass Observation, an architecture that adds read-only observation heads to selected Transformer layers so internal hidden states can be inspected without altering the model's behavior. The approach aims to bridge the gap between the high-dimensional reasoning space of large language models and the text-only outputs users normally see. It is presented as a conceptual design for non-intrusive semantic extraction.

interpretability/llm/research/architecture

papersTODAY 04:00 UTC

arXiv paper proposes scalable data attribution via influence matrix estimation

A new arXiv preprint addresses the computational cost of data attribution, which measures how individual training samples affect a model's behavior. The authors frame the problem around estimating the influence matrix at scale, with applications in data valuation, machine unlearning, and interpretability. The abstract highlights that scaling such methods has remained a longstanding obstacle.

data-attribution/research/interpretability/scalability

papersTODAY 04:00 UTC

Bangla Sentence Function Classification Corpus and Benchmark Released

Researchers present a new annotated corpus for classifying sentence functions in Bangla, a resource previously lacking for the language. The work benchmarks several models on the task and adds interpretability analysis of their predictions. Such sentence-type identification supports dialogue systems, speech synthesis, and machine translation.

nlp/bangla/benchmark/corpus

papersSEP 12 13:27 UTC

Study links reasoning models' internal states to distinct thought steps

A new study finds that operations such as arithmetic, recalling formulas, and logical deduction show up as separate patterns inside reasoning models, most visibly in their middle layers. This suggests models carry out more processing than their published chain-of-thought text discloses, which researchers flag as relevant to AI safety and oversight. The findings could inform how developers monitor or audit model reasoning.

interpretability/reasoning/safety/research

papersSEP 12 04:00 UTC

arXiv paper presents flexible, interpretable method for measuring accent distance

A new arXiv preprint addresses how to quantify the differences between two speakers' accents, a task relevant to both linguistics and speech technology. The authors note that measurement approaches currently differ depending on the research field, such as phonetics, and propose a method intended to be both adaptable and interpretable. The announcement provides only the abstract, so experimental details and results are not yet visible from the listing.

speech/linguistics/accent/research

papersSEP 12 04:00 UTC

Paper Identifies 'Perfect Aliasing' Failure in Compliant-Context Truth Probes

A new arXiv paper examines a problem it calls "perfect aliasing," in which a truthfulness probe trained on data where honest reporting and the task's prescribed action line up cannot tell those two targets apart from the labels alone. The authors argue this amounts to a failure of semantic identification, and they study it using a controlled binary reporting setup.

interpretability/truthfulness/alignment/safety

papersSEP 10 04:00 UTC

Central Dogma Transformer II: an interpretable model for cellular gene regulation

A revised arXiv preprint introduces Central Dogma Transformer II, a transformer-based model framed as an 'AI microscope' for studying cellular regulatory mechanisms. The authors argue that gene-regulation research requires models whose learned internal structure can be directly examined and mapped onto regulatory relationships, rather than opaque but accurate predictors. The update refines the paper's case for interpretability as a core requirement in computational biology.

interpretability/biology/gene-regulation/transformers

papersSEP 10 04:00 UTC

LM-X: Explainable Vision-Language-Action Model Predicts Progress, Events, and Uncertainty

Researchers introduce LM-X, a framework for vision-language-action robot policies that exposes an explanatory state alongside its actions. Instead of acting as a stimulus-to-action black box, the model natively predicts task progress, notable events, and uncertainty in its decisions. The work aims to bring interpretability to large-scale generalist robot control.

robotics/explainable-ai/multimodal

papersSEP 10 04:00 UTC

MonoTM uses sparse autoencoders for interpretable topic modeling features

A new arXiv paper presents MonoTM, a topic modeling method that moves past conventional top-word lists by deriving interpretable features from sparse autoencoders. By decomposing dense text representations into monosemantic components, the approach aims to describe topic semantics in large corpora more faithfully.

topic-modeling/interpretability/sparse-autoencoders/nlp

papersSEP 10 04:00 UTC

Researchers probe whether speech foundation models truly learn words

A new arXiv study investigates self-supervised speech foundation models, which are widely deployed for speech recognition and to supply tokens for speech-capable language models. The authors analyze what these models' internal representations encode, testing whether they capture genuine word-level linguistic structure rather than only acoustic patterns. The results bear on how such models should be interpreted and used in downstream speech applications.

speech/foundation-models/interpretability/representations