…Abstract A 3-billion-parameter Sardinian language model was developed using continued pretraining and supervised fine-tuning on limited computational resources, showing superior performance in translation tasks compared to baseline models. AI…
…no frontier model recovers more than a fraction of its full-information performance when deciding whether to ask. Failure analysis identifies three key help-seeking patterns: overconfident wrong beliefs with no gap…
…Generated by Qwen/Qwen2.5-Coder-32B-Instruct Deep research agents have attracted increasing attention for their ability to collect large-scale online information to acquire target knowledge, with recent efforts shifting…
…Jian Mu , , , , Abstract DRIFT is a framework that combines offline trajectories with importance-weighted supervised fine-tuning to achieve multi-turn interactive learning efficiency and performance comparable to reinforcement learning. Generated by…
…Nguyen , Raghav Mehndiratta , Lindsay Devon Brin , , Hari Subramani , Anil Madamala , Sridhar Krishna Nemala , Abstract EVA-Bench presents a comprehensive evaluation framework for voice agents that simulates realistic conversations and measures performance across…
…It supports structural modifications , mathematical transformations , and tensor reshaping through expressive regex and structural targeting , while built-in assertions validate tensor shapes , data types , and values to prevent silent errors. We envision…
…AI-generated summary The landscape of high-performance image generation models is currently shifting from the inefficient multi-step ones to the efficient few-step counterparts (e.g, Z-Image-Turbo and…
…AI-generated summary Geospatial reasoning requires models to resolve complex spatial semantics and user intent into precise target locations for Earth observation. Recent progress has liberated the reasoning path from manual curation…
…humans can often infer the target evaluation criteria from only a few examples, while models are usually trained on hundreds of thousands of comparisons. We present RewardHarness, a self-evolving agentic reward…
…Xinjie Shen , , , , , , , , Abstract Multi-turn dialogue safety monitoring system detects harmful intent accumulation through turn-level analysis and evaluates performance on a new benchmark dataset. AI-generated summary Hidden malicious intent in…