Tag

OSReward: Instituting Standardized Evaluation for Cross-Platform Computer-Use Reward Models

Evaluates vision-language model judges on computer-using agent (CUA) trajectories to measure verifier reliability. Provides OSReward-Hard and OSReward-Multi challenge sets, the OS-Shepherd-100K reasoning-annotated corpus, and trained OS-Shepherd reward models that match commercial judges at ~30–60× lower cost.

Recursive Task Synthesis

Provides 37,484 validated command-line tasks generated by recursive task synthesis, each paired with searchable metadata and a sanitized, runnable package (instructions, solution, verifier, and optional Dockerfile).

Reinforcement Learning with Evolving Rubrics as Rewards for Audio Reasoning

Supervises audio reasoning by generating per-sample, audio-grounded rubrics that evolve with model rollouts and serve as reinforcement-learning rewards, improving perception and adaptive multi-step reasoning while avoiding reward saturation.

SFT Conflicts, RL Coexists: A Theoretical and Empirical Analysis of Multi-Task Learning for LLMs

Analyzes why supervised fine-tuning (SFT) causes severe task conflicts under multi-stage multi-task training while reinforcement learning (RL) enables stable coexistence, attributing the effect to sparse, near-orthogonal RL parameter updates and proposing Parallel-RL to decouple multi-task training.

Beyond Simply Environment Scaling: Designing Effective Environment Distributions for Multimodal Agent Learning

Analyzes how to build effective training environment distributions for multimodal agents and proposes Ability-aware Environment Selection (AES) and Hierarchical Difficulty Curriculum (HDC) to improve diversity and difficulty scheduling, yielding large relative gains in experiments.

ToolArtist: Tool-Using Unified Multimodal Models for Agentic Image Generation

Orchestrates reasoning, external tool use, and native image generation under one unified multimodal agent policy via post-training. Introduces RAD-GRPO for agentic reinforcement fine-tuning and releases training data plus the full post-training infrastructure.

Nemotron-RL-Agentic-Terminal-Pivot-v1

Provides per-decision training samples for RL-driven command-line LLM agents: each record pairs a task prompt plus terminal history with a teacher's next-action in Terminus-2 JSON. Around 31k verifier-passing samples from 630 ATCB tasks, formatted for NeMo Gym's terminus_judge and licensed CC-BY-4.0.

AgentOPSD: Recursive Self-Distillation for Agentic Reinforcement Learning

Converts sparse trajectory-level rewards into turn-level credit by aggregating token-level teacher–student log-probability gaps and recursively updating a Bayesian belief in log-odds; produces turn-wise reweighting for policy optimization without an extra critic or rollouts.

Learning from Failures: Retrieval-Centric CoT via Hard Negatives for Unified Multimodal Retrieval

Generates retrieval-centric Chain-of-Thought (RC-CoT) over initially retrieved candidates to improve unified multimodal retrieval via reranking or full-corpus re-retrieval with a dual-mode embedder. Trains an embedder–adviser framework (UniME-R1) using mined hard negatives, supervised learning, and retrieval-oriented reinforcement learning.

SimWAM: A Simple World Action Model for End-to-End Autonomous Driving

Uses video generation only as a training signal to co-train a pretrained video expert and a lightweight action expert, then discards the video branch at inference to produce a low-latency end-to-end driving planner; enhanced with RL for compositional driving rewards.

Macaron-V1: Towards Open Continual Learning with Self-Improvement and Mixture-of-LoRA

A research report proposing a continual-learning agent workflow that pairs recursive self-improvement with a Mixture-of-LoRA design: freeze a foundation model, compose specialist LoRA adapters routed per user turn, and support them with long-context RL and post-training infrastructure.

Intern-S2-Preview: Scientific Agentic Foundation Model

Supports multimodal scientific understanding, long-horizon agentic workflows and scientific tool interaction using a unified pipeline of multimodal pretraining, supervised fine-tuning and scalable multi-task reinforcement learning. Distinctive features include time-series modules for signal forecasting and a separate Memory Decoder that enables rapid domain specialization without changing the frozen 397B backbone.