Notizie
Trusted Uncertainty in Large Language Models: A Unified Framework for Confidence Calibration and Risk-Controlled Refusal
arXiv:2509.01455v1 Announce Type: new Abstract: Deployed language models must decide not only what to answer...
Truncated Step-Level Sampling with Process Rewards for Retrieval-Augmented Reasoning
arXiv:2602.23440v3 Announce Type: replace Abstract: Reinforcement learning has emerged as an effective paradigm for training...
Trump’s mass firing just dealt another blow to American science
This past week delivered another gut punch for science in the US. This time, the...
Trump’s AI protectionism has come for robotics
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories...
Trinity-RFT: A General-Purpose and Unified Framework for Reinforcement Fine-Tuning of Large Language Models
arXiv:2505.17826v2 Announce Type: replace-cross Abstract: Trinity-RFT is a general-purpose, unified and easy-to-use framework designed for...
TRIM: Token-wise Attention-Derived Saliency for Data-Efficient Instruction Tuning
arXiv:2510.07118v2 Announce Type: replace Abstract: Instruction tuning is essential for aligning large language models (LLMs)...
Treating enterprise AI as an operating layer
There’s a fault line running through enterprise AI, and it’s not the one getting the...
Transforming Wearable Data into Personal Health Insights using Large Language Model Agents
arXiv:2406.06464v3 Announce Type: replace-cross Abstract: Deriving personalized insights from popular wearable trackers requires complex numerical...
Transferring Expert Cognitive Models to Social Robots via Agentic Concept Bottleneck Models
arXiv:2508.03998v1 Announce Type: new Abstract: Successful group meetings, such as those implemented in group behavioral-change...
TransEvalnia: A Prompting-Based System for Fine-Grained, Human-Aligned Translation Evaluation Using LLMs
Translation systems powered by LLMs have become so advanced that they can outperform human translators...
Trajectory Releases a Concurrent Multi-LoRA Training Stack for Continual Learning, Reporting a 2.81× Experiment-Throughput Gain
Trajectory’s concurrent multi-LoRA stack reports a 2.81× experiment-throughput gain over single-tenant RL, with all code...
Traits Run Deep: Enhancing Personality Assessment via Psychology-Guided LLM Representations and Multimodal Apparent Behaviors
arXiv:2507.22367v1 Announce Type: new Abstract: Accurate and reliable personality assessment plays a vital role in...
