This paper introduces PraMem, a method that constructs experiential memory by practicing on long historical sequences to address challenges in long-horizon behavior prediction with LLMs, achieving superior performance.

Proposes a multimodal framework that jointly improves automatic speech recognition (ASR) and dialect identification (DID) for Indian languages. Achieves 81.63% DID accuracy, 4.65% CER, and 17.73% WER on 8 languages with 33 dialects.

As LLMs are increasingly deployed as autonomous adjudicators in semi-open textual game environments, robust rule adherence becomes critical when user intent conflicts with system rules. However, these models are trained to be helpful and compliant, leaving them vulnerable to a class of attacks we term Rhetorical Injection, where adversarial users exploit narrative framing techniques such as pseudo-logical reasoning and authoritative coercion to bypass adjudication logic. We present CoC-Seduce, a multi-agent adversarial benchmark built on Tabletop Role-Playing Game (TRPG) mechanics, an ideal instantiation of semi-open environments where rules are explicit for adjudication, yet interaction remains entirely in natural language. Three frontier models, i.e., GPT-5.4, Claude Sonnet 4.6, Gemini…

Gemma 4, the latest generation of open-weight, natively multimodal language models in the Gemma family, featuring dense and Mixture-of-Experts architectures ranging from 2.3B to 31B parameters. It includes improved vision/audio encoders, a unified encoder-free architecture for the 12B model, and a thinking mode that generates reasoning traces before responding. Enhancements in inference speed, memory efficiency, and long-context capabilities lead to strong performance on STEM, multimodal, and long-context benchmarks, rivaling larger frontier open models.

A new approach uses text-to-speech (TTS) to generate training data for Spoken Question Answering in Luxembourgish, a low-resource language. By translating existing text QA pairs, synthesizing speech, and training a parameter-efficient model with a frozen Whisper encoder and LoRA adapters, the method outperforms single-source TTS. Multi-source synthetic data and voice design improve performance, and no-reference TTS quality scores don't reliably predict QA accuracy.

Researchers propose a reinforcement learning with verifiable rewards (RLVR) method to adapt audio-language models for code-switched automatic speech recognition. Using only 10% of the data, RLVR matches the performance of full-dataset supervised fine-tuning on Qwen2-Audio across 10 language pairs, with gains transferring zero-shot to human-recorded speech.

This paper presents a multilingual multimodal NLP framework integrating XLM-RoBERTa, CLIP, multi-head attention, sarcasm, and geospatial metadata to detect misinformation and violence-prone dynamics early. Using a fused dataset of 138,256 Bangla and English samples, it achieves 98% test accuracy with strong precision and recall.

This paper introduces FCPA, a training objective that aligns validator and generator via frequency-corrected consistency, achieving up to +27pp Pearson correlation gain on IFEval and HumanEval.

This paper investigates whether risk aversion trained in low-stakes scenarios can generalize to astronomically high-stakes scenarios, as a potential failsafe against AI misalignment. Introducing the RiskAverseOOD benchmark, initial experiments on Qwen3-8B show that learned risk aversion can partially generalize across 98 orders of magnitude, boosting cooperation rates from 2% to 70% (SFT), 52% (DPO), and 39% (activation steering). However, consistency is insufficient for a reliable failsafe.

Predicting thermal behavior in high-performance EV powertrains is challenging due to unobservable internal temperatures and domain shift from lab to track. This paper applies conformal prediction with weighted ensemble batch prediction intervals (EnbPI) to improve coverage under covariate shift. The method recovers coverage from 70.13% to 72.42% on real battery data, and is tested on Formula 1 telemetry as an unsupervised diagnostic.

LiNO is a multiresolution neural operator built on the second-generation wavelet lifting scheme, learning data-driven multiscale decomposition and evolving coarse and directional detail coefficients separately for scale-aware physics modeling. It outperforms state-of-the-art neural operators on benchmarks including Darcy flow, Poisson equation, Allen-Cahn equation, compressible Navier-Stokes equation, and Gray-Scott reaction-diffusion system.

This study develops a granularity-aware EEG feature pipeline that organizes multi-scale descriptors into global, regional, and channel levels. Using the HBN cohort, it predicts four psychopathology dimensions. Tree-based models and granularity-balanced feature selection show modest improvements. Visualization reveals dimension-specific patterns. A cross-dataset check on PEARL suggests technical feasibility.

A new generator-agnostic method selects the most informative synthetic images from a fixed pool by splitting classes into homogeneous and heterogeneous subsets, improving downstream task performance with up to 40% fewer samples.

This paper applies Federated Learning (FL) to object detection in drone networks, allowing drones to collaboratively train a shared model without sharing raw aerial imagery. Using the Sherpa.ai FL platform on the KIIT-MiTA dataset, the authors compare FL with single-drone and centralized baselines. Their best lightweight model (YOLO26 nano) achieves relative gains of 52.89% in [email protected] and 67.80% in [email protected]:0.95 over single-drone training, while remaining close to centralized performance. The results demonstrate that FL enables scalable, high-performing, and privacy-preserving object detection across distributed drone fleets.

GRAFT is a per-word pronunciation conditioning mechanism for text-to-speech neural codec language modeling. It controls pronunciation of a chosen word from a short spoken sample, using voice conversion to decouple hint speaker from target speaker. In a blind English listening study, human raters rank GRAFT first, and on a five-language benchmark, it reduced target-word phoneme error rate by 22-39% while preserving speaker similarity and naturalness.

QuantFlow is a probabilistic forecasting framework that combines inverted sequence embedding, bidirectional Mamba state-space decoders, quantile regression, and federated learning. Experiments demonstrate strong performance across multiple datasets and effective accuracy retention in non-IID federated settings, while also revealing limitations on irregular epidemiological signals and long-horizon generalization.

This study proposes a two-dataset benchmarking framework to fairly evaluate time series foundation models (TSFMs) for electricity price forecasting. It finds TSFMs competitive but dependent on covariate support, not always surpassing domain-specific methods. Ensembles show promise, capturing complementary information.

This paper argues that perturbation-based construct-validity audits are fragile: their conclusions can be silently manufactured by implementation details invisible to readers. It names five pipeline failure modes (F1–F5) and demonstrates them via a self-audit on safety benchmarks and open-weight instruction-tuned models. Under a unified six-point due-diligence gate, every cell lands in a non-confirmatory bucket. The gate is positioned as a withholding and disclosure protocol for assurance-grade evidence, supplementary to classical construct-validity evidence.

This paper proposes a sliding-window-based reinforcement learning (SWRL) framework for end-to-end online scheduling in flexible assembly flow shop scheduling with complex kitting constraints. It formulates the problem as a heterogeneous graph Markov decision process, and integrates sliding-window filtering, spatiotemporal graph encoding, and dynamic action mapping. Experiments on real-world home appliance manufacturing data show that SWRL achieves consistent tardiness reductions over classical dispatching rules and existing deep reinforcement learning methods, with robust performance across varying resource configurations.

VERITAS is a domain-agnostic replication framework built around CLI coding agents. It extracts claims from papers, runs methodologies while resolving issues, and judges claims against experimental evidence. Evaluated on 65 papers across multiple disciplines, VERITAS achieves state-of-the-art performance on CORE-Bench and ReplicationBench.

Large language models (LLMs) face a persistent challenge in balancing safety, helpfulness, and trustworthiness. Traditional refusal-oriented alignment strategies mitigate harmful content but often fail to serve legitimate user needs. Oyster-II proposes a reinforcement learning (RL)-based constructive safety alignment framework, employing a Zero-RL paradigm combined with a multi-stage RL strategy. It addresses two critical limitations of Oyster-I's Supervised Fine-Tuning (SFT) scheme: insufficient safety generalization to out-of-distribution scenarios and safety chain-of-thought (CoT) over-generalization. Evaluated on extensive benchmarks, Oyster-II comprehensively surpasses Qwen3-14B and Oyster-I on safety dimensions, achieving cross-scale performance comparable to Qwen3-Max and Qwen3.5-3…

A new benchmark MedCalc-Pro evaluates LLMs on complex medical calculations involving multiple calculators, nested scales, and fuzzy queries. The authors also propose a generalizable agent framework that outperforms existing models on all three task settings.

Large language model (LLM) agents can improve through accumulated experience, but free-form textual memories become difficult to maintain, validate, and reuse as interactions grow. This paper proposes Object-Centric Environment Modeling (OCM), which organizes experience into an executable object-centric environment model. OCM maintains two connected code bases: object knowledge (defining environment entities as Python classes) and procedure knowledge (recording reusable interaction patterns). Experiments show OCM achieves the best average rank across benchmarks and reduces invalid actions.

SwarmResearch is an orchestrator-subagent harness that uses global context to steer a population of search agents, achieving better or comparable results on open-ended optimization tasks compared to state-of-the-art methods.

Researchers introduce REDI, an open-source framework that automates the transformation of large-scale scientific datasets into AI-ready data through a unified five-stage pipeline. It includes provenance tracking, reproducibility, and agent-native deployment. Tested on climate, proteomics, materials science, and nuclear fusion, it shows near-ideal parallel scaling and identifies file I/O as the primary cost.

Reinforcement learning agents under partial observability can benefit from SLM guidance, but vanilla uncertainty-gating fails. The proposed ASK+ provides trajectory-aware context and chain-of-thought reasoning, turning the SLM into an informative consultant. Experiments show significant gains on DoorKey, FourRooms, and HigherLower, with prompt design dominating model scale.

Pairwise comparisons are a common tool for learning preferences, but they assume local comparisons are sufficient and people can answer decisively. This paper examines how internal pluralism—where individuals evaluate rules based on multiple priorities—compromises these assumptions. It identifies failures: global priorities like proportionality may be missed, and tension between priorities can distort behavior. Allowing indecision reduces queries needed, pointing to methods that elicit priorities directly.

iFLYTEK-Embodied-Omni is a unified multimodal foundation model that jointly models vision, language, and action. It employs a brain-cerebellum architecture: a vision-language model and video generation model serve as the high-level brain for understanding and planning, while an action generation model acts as the low-level cerebellum for executing actions. The model is trained via a four-stage strategy on a comprehensive dataset combining human demonstrations, robot interactions, and general image-text data.

Social media is rebranding surveillance technology like AI glasses as fashionable, normalizing harassment and non-consensual filming. The article uses Alexa Chung's Ring camera outfit posts as an example to highlight this troubling trend.

Universities increasingly use AI-detection tools to identify AI-generated student work, but studies show high false-positive rates, including false flags on human-written texts like the US Declaration of Independence, raising fairness concerns.