AI News Archive: May 21, 2026 — Part 18
Sourced from 500+ daily AI sources, scored by relevance.
- Chinese sensorimotor and embodiment norms for 3,000 lexicalized concepts
Understanding how conceptual knowledge is grounded in bodily experience, and to what extent machine systems can acquire such knowledge without direct sensorimotor experience, are central questions in both cognitive science and embodied artificial intelligence research. Large-scale normative resource...
- Agentic CLEAR: Automating Multi-Level Evaluation of LLM Agents
Agentic systems are becoming more capable: agents define strategies, take actions, and interact with different environments. This autonomy poses serious challenges for overseeing and assessing agent behavior. Most current tools are limited, focusing on observability with basic evaluation capabilitie...
- Beyond Temperature: Hyperfitting as a Late-Stage Geometric Expansion
Recent work has identified a counterintuitive phenomenon termed "Hyperfitting", where fine-tuning Large Language Models (LLMs) to near-zero training loss on small datasets surprisingly enhances open-ended generation quality and mitigates repetition in greedy decoding. While effective, the underlying...
- LANG: Reinforcement Learning for Multilingual Reasoning with Language-Adaptive Hint Guidance
Reinforcement learning has proven effective for enhancing multi-step reasoning in large language models (LLMs), yet its benefits have not fully translated to multilingual contexts. Existing methods struggle with a fundamental trade-off: prioritizing input-language consistency severely hampers reason...
- SpaceDG: Benchmarking Spatial Intelligence under Visual Degradation
Multimodal Large Language Models (MLLMs) have made rapid progress in spatial intelligence, yet existing spatial reasoning benchmarks largely assume pristine visual inputs and overlook the degradations that commonly occur in real-world deployment, such as motion blur, low light, adverse weather, lens...
- Search-E1: Self-Distillation Drives Self-Evolution in Search-Augmented Reasoning
Post-training has become the dominant recipe for turning a language model into a competent search-augmented reasoning agent. A line of recent work pushes its performance further by adding elaborate machinery on top of this standard pipeline. These augmentations import external supervision from stron...
- BeLink: Biomedical Entity Linking Meets Generative Re-Ranking
Despite recent progress, Biomedical Entity Linking (BEL) with large language models (LLMs) remains computationally inefficient and challenging to deploy in practical settings. In this work, we demonstrate that instruction-tuning of open-source generative models can offer an effective solution when a...
- In Silico Modeling of the RAMPHO Buffer: Dissociating Informational and Energetic Masking via Phonetic Entropy in Deep Neural Networks
The fundamental challenge of listening in multi-talker environments is a cognitive bottleneck, defined by the Ease of Language Understanding (ELU) model as a failure within the RAMPHO episodic buffer. Current deep neural networks for speech enhancement optimize purely for physical acoustics, failing...
- From Correlation to Cause: A Five-Stage Methodology for Feature Analysis in Transformer Language Models
We propose a five-stage methodology for causal feature analysis in transformer language models (probe design, feature extraction, causal validation, robustness testing, and deployment integration) and demonstrate it end-to-end on GPT-2 small performing the Indirect Object Identification (IOI) task. ...
- Assisted Counterspeech Writing at the Crossroads of Hate Speech and Misinformation
Hate speech and misinformation frequently co-occur online, amplifying prejudice and polarization. Given their scale, using Large Language Models (LLMs) to assist expert counterspeech (CS) writing has gained interest, yet prior work has addressed these phenomena separately. We bridge this gap by stud...
- DeferMem: Query-Time Evidence Distillation via Reinforcement Learning for Long-Term Memory QA
Large language model (LLM) agents still struggle with long-term memory question answering, where answer-supporting evidence is often scattered across long conversational histories and buried in substantial irrelevant content. Existing memory systems typically process memory before future queries are...
- Unified Data Selection for LLM Reasoning
Effectively training Large Language Models (LLMs) for complex, long-CoT reasoning is often bottlenecked by the need for massive high-quality reasoning data. Existing methods are either computationally expensive or fail to reliably distinguish high- from low-quality reasoning samples. To address this...
- Modeling Pathology-Like Behavioral Patterns in Language Models Through Behavioral Fine-Tuning
Large language models are increasingly used as computational tools for modeling human-like behavior. We introduce a behavioral induction framework that modifies model policies through fine-tuning on structured decision-making tasks: using synthetic datasets inspired by maladaptive behavioral pattern...
- Harder to Defend: Towards Chinese Toxicity Attacks via Implicit Enhancement and Obfuscation Rewriting
Large language models (LLMs) require robust toxicity evaluation beyond explicit wording. This setting remains underexplored in Chinese, where toxicity may combine semantic indirectness with surface obfuscation. We introduce Chinese Implicit Toxicity Attack (CITA), a controlled red-team evaluation an...
- Survive or Collapse: The Asymmetric Roles of Data Gating and Reward Grounding in Self-Play RL
Self-play reinforcement learning trains language models on their own generated tasks, co-evolving a proposer and solver without human labels. Recent systems report strong reasoning gains, but collapse and instability are widely observed and poorly understood. The dominant response treats this as a r...
- Evaluation of Chunking Strategies for Effective Text Embedding in Low-Resource Language on Agricultural Documents
In this study, we compare the performance of four text chunking approaches: Recursive, Khmer-Aware, Sentence-Based, and LLM-Based within a Retrieval-Augmented Generation (RAG) framework applied to Khmer agricultural documents. The document chunks are encoded using the BGE-M3 multilingual embedding m...
- Evaluating Commercial AI Chatbots as News Intermediaries
AI chatbots are rapidly shaping how people encounter the news, yet no prior study has systematically measured how accurately these systems, with their proprietary search integrations and retrieval-synthesis pipelines, handle emerging facts across languages and regions. We present a 14-day (February ...
- AnyMo: Geometry-Aware Setup-Agnostic Modeling of Human Motion in the Wild
As wearable and mobile devices become increasingly embedded in daily life, they offer a practical way to continuously sense human motion in the wild. But inertial signals are highly dependent on the sensing setup, including body location, mounting position, sensor orientation, device hardware, and s...
- Tokenization with Split Trees
We introduce Tokenization with Split Trees (ToaST), a subword tokenization method that directly optimizes compression under a new recursive inference procedure. ToaST greedily splits each pretoken into a full binary tree using precomputed byte n-gram counts, independent of any vocabulary. Given a vo...
- Boiling the Frog: A Multi-Turn Benchmark for Agentic Safety
Background. Traditional safety benchmarks for language models evaluate generated text: whether a model outputs toxic language, reproduces bias, or follows harmful instructions. When models are deployed as agents, the safety-relevant object shifts from what the system says to what it does within an e...
- The Double Dilemma in Multi-Task Radiology Report Generation: A Gradient Dynamics Analysis and Solution
While multi-task learning based automatic radiology report generation (RRG) is widely adopted to ensure clinical consistency, most focus on architectural designs yet remain limited to coarse linear scalarization strategies. These strategies cannot effectively balance the hard constraints of discrimi...
- A Tutorial on Diffusion Theory: From Differential Equations to Diffusion Models
This tutorial develops diffusion models from the viewpoint of differential equations. We begin with the conditional Gaussian forward process and show that this path admits both an ordinary differential equation (ODE) representation and a stochastic differential equation (SDE) representation. Averagi...
- SynAE: A Framework for Measuring the Quality of Synthetic Data for Tool-Calling Agent Evaluations
Today, tool-calling agents are commonly evaluated or tested on static datasets of execution traces, including input commands, agent responses, and associated tool calls. However, internal production datasets are often insufficient or unusable for testing; for example, they may contain sensitive or p...
- One prompt is not enough: Instruction Sensitivity Undermines Embedding Model Evaluation
Instruction embedding models have become common among state-of-the-art models, however are evaluated using a single prompt per task. The single-point evaluation ignores a main problem of the instruction-based approach namely: sensitivity to the phrasing of the instruction. We present an empirical st...
- Reflecti-Mate: A Conversational Agent for Adaptive Decision-Making Support Through System 1 and System 2 Thinking
Making high-stakes personal decisions involves cognitive, emotional, and intuitive processes, and individuals differ in how they allocate attention across these modes. Integration of these processes has shown to benefit decision making. Yet, most current decision-support systems focus primarily on s...
- Polite on the Surface, Wrong in Practice: A Curated Dataset for Fixing Honorific Failures in Multilingual Bangla Generation
Recent advances in Multilingual Large Language Models (MLLMs) have significantly enhanced cross-lingual conversational capabilities, yet modeling culturally nuanced and context-dependent communication remains a critical bottleneck. Specifically, existing state-of-the-art models exhibit a severe prag...
- Structured-Sparse Attention for Entity Tracking with Subquadratic Sequence Complexity
Entity tracking requires maintaining and updating latent states for entities and attributes over long sequences. Recent task-specific attention operators can compress deep Transformer stacks into a few layers by performing multi-hop state propagation within a single layer, but their dense evaluation...
- Epicure: Navigating the Emergent Geometry of Food Ingredient Embeddings
We present Epicure, a family of three sibling skip-gram ingredient embeddings retrained from scratch on a multilingual recipe corpus. We aggregate 4.14M recipes from 11 sources spanning seven languages, English, Chinese, Russian, Vietnamese, Spanish, Turkish, Indonesian, German, and Indian-English, ...
- Slideshot
Product demo videos, recorded by your AI agent
- Boundary-targeted Membership Inference Attacks on Safety Classifiers
Safety classifiers are essential safeguards within generative AI systems, filtering harmful content or identifying at-risk users when interacting with large language models. Despite their necessity, these models are trained on sensitive datasets including discussions of self-harm and mental health, ...
- TransitLM: A Large-Scale Dataset and Benchmark for Map-Free Transit Route Generation
Public transit route planning traditionally depends on structured map infrastructure and complex routing engines, and no existing dataset supports training models to bypass this dependency. We present TransitLM, a large-scale dataset of over 13 million transit route planning records from four Chines...
- GHI: Graphormer over Conditioned Hypergraph Incidence for Aspect-Based Sentiment Analysis
Aspect-based sentiment analysis (ABSA) requires models to bind sentiment evidence to the correct aspect, making it a natural testbed for fine-grained structural reasoning. We introduce GHI, a Graphormer-over-Conditioned-Hypergraph-Incidence framework that is designed as an incidence-based structural...
- Which Way Did It Move? Diagnosing and Overcoming Directional Motion Blindness in Video-LLMs
Video Large Language Models (Video-LLMs) have made rapid progress on temporal video understanding, yet many fail at a basic perceptual primitive: signed image-plane motion direction. On simple videos of a single object moving left, right, up, or down, most Video-LLMs perform near chance, with above-...
- Cambrian-P: Pose-Grounded Video Understanding
Camera pose matters. The position and orientation of each viewpoint define a shared spatial coordinate frame that relates observations across video frames. Yet this signal is largely absent from multimodal LLMs (MLLMs) for video understanding, which process frames as isolated 2D snapshots, instead o...
- Sensor2Sensor: Cross-Embodiment Sensor Conversion for Autonomous Driving
Robust training and validation of Autonomous Driving Systems (ADS) require massive, diverse datasets. Proprietary data collected by Autonomous Vehicle (AV) fleets, while high-fidelity, are limited in scale, diversity of sensor configurations, as well as geographic and long-tail-behavioral coverage. ...
- Synthetic Data Alone is Enough? Rethinking Data Scarcity in Pediatric Rare Disease Recognition
Children with rare genetic diseases often exhibit distinctive facial phenotypes, yet developing computer vision systems for early diagnosis remains challenging due to extreme data scarcity, privacy constraints, and limited data sharing in pediatric settings. These challenges not only hinder automate...
- Conceptualizing Embeddings: Sparse Disentanglement for Vision-Language Models
Vision-language models learn powerful multimodal embeddings, yet their internal semantics remain opaque. While sparse autoencoders (SAEs) can extract interpretable features, they rely on expanding the representation dimension, which compromises the original geometry and introduces redundancy. We int...
- What Does the Caption Really Say? Counterfactual Phrase Intervention for Compositional Data Selection in Vision-Language Pretraining
CLIP-style contrastive pretraining typically curates web-scale image-text pairs using sample-level filtering signals, often based on pair-level alignment. We show that this signal saturates: once coarse mismatches are removed, stricter global filtering no longer tracks the compositional supervision ...
- H-Flow: Self-supervised Human Scene Flow via Physics-inspired Joint Multi-modal Learning
Parametric human models capture global pose but cannot represent the non-rigid surface dynamics of clothing and soft tissue. Generic scene flow estimates dense motion but breaks down on articulated bodies, where pixel-level supervision is also intractable to acquire. We introduce H-Flow, a dense hum...
- Decoupling Ego-Motion from Target Dynamics via Dual-Interval Motion Cues for UAV Detection
Object detection from Unmanned Aerial Vehicles (UAVs) is challenged by severe ego-motion, camera jitter, and large scale variations. While modern detectors perform well on static images, their direct application to UAV video often fails, particularly for small objects in dynamic scenes. Existing mot...
- Beyond Chamfer Distance: Granular Order-aware Evaluation Metric For Online Mapping
Online map estimation is a crucial component of autonomous driving systems that reduces the reliance on costly high-definition maps. State-of-the-art (SOTA) methods commonly predict map elements as ordered sequences of points that form polylines and polygons. The evaluation of these methods relies p...
- VGenST-Bench: A Benchmark for Spatio-Temporal Reasoning via Active Video Synthesis
Spatio-temporal reasoning is a core capability for Multimodal Large Language Models (MLLMs) operating in the real world. As such, evaluating it precisely has become an essential challenge. However, existing spatio-temporal reasoning benchmark datasets primarily rely on static image sets or passively...
- Cell Phantom Video Generation in Elliptical Fourier Descriptor Domain
Training Deep Neural Networks for tracking individual cells in biomedical videos requires a large amount of annotated data. The annotation of videos for cell tracking is very time consuming and often requires domain expertise; this explains the limited availability of public annotated data to addres...
- GeoWeaver: Grounding Visual Tokens with Geometric Evidence before Scene Reasoning
Spatio-temporal reasoning in vision-language models requires visual representations that preserve physical geometry rather than merely semantic appearance. Recent multimodal models incorporate geometric information through structural branches, 3D-aware supervision, reasoning-stage fusion, or long-ho...
- Visual Usability Checker
Validate your design decisions instantly with AI insights
- MOTOR: A Multimodal Dataset for Two-Wheeler Rider Behavior Understanding
Two-wheelers account for a disproportionately high share of road fatalities in the Global South. Research on two-wheeler rider behavior, however, lags far behind four-wheelers, where multimodal datasets have driven major advances in Advanced Driver Assistance Systems (ADAS). To address this gap, we ...
- Case-Aware Medical Image Classification with Multimodal Knowledge Graphs and Reliability-Guided Refinement
Deep learning has brought significant progress to medical image classification, yet most existing methods still rely on isolated visual evidence and cannot effectively leverage similar cases or external knowledge. In clinical practice, diagnosis is typically supported by historical similar cases and...
- Predicting 30-Day Heart Failure Readmissions Using Machine Learning: Insights From the Kansas Health Information Network (KHIN)
Background: Heart failure (HF) is a major contributor to inpatient hospital utilization, with persistently high 30-day readmission rates. Existing prediction tools are frequently restricted to primary-diagnosis HF admissions, potentially excluding clinically relevant HF-related hospitalizations. Objectives: To develop and validate risk prediction models using machine learning (ML)-based risk prediction models to predict 30-day readmissions among patients with HF using the Kansas Health Information Network, a statewide health information exchange. Methods: This retrospective cohort study analyzed HF hospitalizations using predictors including demographics, comorbidities, laboratory results, medications, clinical quality metrics for diabetes and kidney disease management, and prior healthcare utilization. Five ML models, including regularized logistic regression, random forest, extreme gradient boosting, categorical boosting, and deep neural network, were trained using stratified 5-fold
- Interpretable Symptom-Based Machine Learning for Parkinson's Disease Prediction: A Feasibility Study
Background: Parkinson's disease (PD) has a prolonged prodromal phase during which non-motor symptoms (NMS) may emerge years before the appearance of classical motor signs. This makes NMS a promising and clinically accessible source of information for early risk stratification. Objective: In this study, we investigated whether NMS alone can serve as reliable predictors of PD risk using clinical data from the Parkinson's Progression Markers Initiative (PPMI) cohort. Methods: We developed a stacked ensemble machine learning framework that integrates feature-level modelling, a global multivariate model, and a patient-similarity component to capture complementary patterns within NMS profiles. The model was trained using leakage-controlled patient-level validation and evaluated on an independent held-out test set. Results: The final ensemble achieved strong predictive performance, with an area under the ROC curve of 0.955, sensitivity of 0.929, and specificity of 0.900. Explainability analys
- Automated Macrolinguistic Discourse Analysis for Transdiagnostic Detection of Language Impairments
Macrolinguistic discourse analysis offers valuable insight into how patients with neurogenic communication disorders organize and produce informative speech, yet it remains a largely manual and labor-intensive process. We report an automated pipeline for macrolinguistic discourse analysis for individuals with aphasia and dementia that integrates automatic speech recognition (ASR), utterance segmentation, sentence-level embeddings, centroid-based main-concept matching, and rule-based coherence error classification. These algorithms were applied to Cinderella story retellings from 309 participants (113 controls, 102 post-stroke aphasia (PWA), and 94 dementia). The algorithm reliably identified main concepts (83% accuracy against human labels) and derived interpretable features such as semantic distance to a main concept centroid, main concept coverage, and coherence error rates. Crucially, diagnostic classification results showed that logistic-regression classifiers trained on 10 macroli