Category
4,653 papers
An overview of the ImageEval 2026 shared task on culturally grounded Arabic multimodal evaluation is presented, covering spoken visual question answering and image-grounded hallucination detection in English and Modern Standard Arabic, and CRAI-Bench, evaluating the cultural accuracy of text-to-image generation.
A mechanistic intervention framework for identifying and transferring task-relevant sparse latent features across languages and reframes some cross-lingual reasoning gaps as failures of mechanism elicitation rather than capability absence, and offers a causally testable route to feature-mediated transfer without translation, fine-tuning, or changing the user-facing language.
SemPOI-RL is proposed, a framework that aligns LLM semantic reasoning with structured sequence generation for interpretable OOT recommendation and consistently outperforms both traditional recommenders and direct LLM baselines, while providing interpretable style attribution across different phases of a trip.
This work proposes AutoTraceGT (Automated Trace analysis through Grounded Theory), the first multi-agent pipeline that automates grounded theory on agent trajectories and suggests Grounded Theory offers a scalable analytic tool for ML researchers and agent developers studying what agents actually do.
Reach audiences
Advertise in front of researchers, engineers, and readers.
A fine-tuning-based Stratified Consistency Distillation approach that shows significant and consistent improvements in both Pass@K and the novel Equivalent Logical Similarity metrics, demonstrating the potential of advancing logical translation through consistency distillation.
Logical reasoning with large language models (LLMs) is a critical capability, as it reflects a system's ability to correctly deduce hypotheses from a given context using faithful deductive processes. However, LLM reasoning has often been shown to be sensitive to small surface-level variations in problem formulation, raising questions about whether models truly follow the underlying logical structure. Studying this behavior is challenging because the symbolic components of logical problems, such as operators and predicates, are difficult to systematically manipulate in natural language. We introduce a tool-driven framework for generating controlled, label-preserving edits to logical reasoning problems. Our method operates on symbolic representations of first-order logic and constraint satisfaction problem tasks, enabling targeted modifications to logical operators and other structural components before translating them back into natural language. Using this framework, we evaluate various LLMs under cumulative and individual operator edits and analyze their behavior in response to these changes. Our quantitative and qualitative analyses show that LLM reasoning behavior under controlled operator edits is inconsistent, regardless of model size or family: models sometimes adapt correctly to structural changes but often fail to track their logical consequences. The results from this automated stress test enable an evaluation of language models across different dimensions and help measure the reliability of their reasoning.
This work introduces the Differential Reasoning Router (DRR), a cost-aware framework for cold-start LLM annotation that jointly optimizes model selection and human escalation, enabling a gradual shift from human-heavy cold-start annotation toward high-confidence automated routing.
A Label-Guided Neural Topic Model (LGNTM) is proposed, which learns dedicated label-aligned topics, grounds them in lexical and document semantic spaces, and preserves consistency between topic structures and label structures.
This work proposes CPR (Critical-Point Routing), a token-level routing framework between a base model and its expert derivative, based on critical tokens where the base model fails but the expert succeeds, and achieves state-of-the-art across all settings.
Over six months with four state-of-the-art LLM agents configured with web search, aggregate nowcast accuracy is broadly comparable to the institutional and professional benchmarks, with performance varying widely across individual indicators.
AtlasNLP is introduced, a country-aware atlas of over 13,000 NLP dataset records across normalized NLP task categories, tracking both the populations represented and where datasets are produced, showing that dataset coverage is highly uneven across countries and tasks and language coverage does not imply geographic representation.
We present Arkios, a 1.04B-parameter dense transformer pretrained from scratch on 150B tokens of bilingual English-Nepali text, using a custom single-file C/CUDA training stack and a Devanagari-aware byte-level BPE tokenizer built for this project. On ARC-Easy and ARC-Challenge, Arkios exceeds three comparably sized open models (Pythia-1.4B, TinyLlama-1.1B, OLMo-1B) despite an order of magnitude fewer training tokens, likely aided by a match between our educational-web-text pretraining data and ARC's grade-school-science format rather than a general capability advantage. We report full evaluation results under standard protocols, including a correction to an earlier partial-sample estimate, and findings specific to evaluating small models in a low-resource language: the standard multiple-choice-letter prompt format used by common evaluation harnesses places this model at chance on Nepali reading comprehension, and simultaneously at chance on English in the same format, which would lead a naive benchmark run to conclude the model has no Nepali ability when in fact it does. Concretely, both languages score at chance in the letter-choice format (0.240 Nepali, 0.236 English, against a chance baseline of 0.250), while scoring the answer text directly reveals genuine, English-favoring comprehension (0.306 Nepali, 0.387 English). We describe a manifest-conditioned tool-use contract introduced during instruction tuning, where tool calls are permitted only when a tool manifest is declared in context and suppressed otherwise, and report where that contract holds and where it does not. We release both the base and instruction-tuned model weights under Apache-2.0. The training code and a small privately-sourced portion of the Nepali pretraining corpus are not released; everything needed to reproduce the reported numbers from the released weights is included here.
Machine Intelligence
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
