DeltaTTG: Training Multi-Turn Agents with a Token-to-Go Progress Meter

cs.LG, cs.AI, cs.CL

Submitted: 2026-10-13

Updated: 2026-10-13

Code: https://github.com/Xuyan923r/DeltaTTG

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Sources

• $\tau^2$-Bench: Evaluating Conversational Agents in a Dual-Control Environment
• BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agent
• Process Reward Models for LLM Agents: Practical Framework and Directions
• Agent-RLVR: Training Software Engineering Agents via Guidance and Environment Rewards
• Inducing Process Supervision from Outcome-Only Reinforcement Learning
• AgentProcessBench: Diagnosing Step-Level Process Quality in Tool-Using Agents
• DARC: Decoupled Asymmetric Reasoning Curriculum for LLM Evolution
• Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning
• VinePPO: Refining Credit Assignment in RL Training of LLMs
• Correct Answers from Sound Reasoning: Verifiable Process Supervision for Language Models
• VICT: Verifier-Instrumented Credit Tracing for Long-Horizon LLM Agent Reinforcement Learning
• Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening
• BAGEN: Are LLM Agents Budget-Aware?
• ABSeeker: Training Long-Horizon Search Agents via Answer-Backtracked Credit Assignment
• Agent Lightning: Train ANY AI Agents with Reinforcement Learning
• Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• HybridFlow: A Flexible and Efficient RLHF Framework
• ALFWorld: Aligning Text and Embodied Environments for Interactive Learning
• Voyager: An Open-Ended Embodied Agent with Large Language Models

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks