MCPDisCo: Seesaw-Mitigated Tuning for MCP-Native Agents
cs.AI
Submitted: 2026-10-10
Updated: 2026-10-10
Sources
- MCP-Atlas: A Large-Scale Benchmark for Tool-Use Competency with Real MCP Servers
- RLFactory: A Plug-and-Play Reinforcement Learning Post-Training Framework for LLM Multi-Turn Tool-Use
- MCP-Zero: Active Tool Discovery for Autonomous LLM Agents
- Scaling Agentic Capabilities, Not Context: Efficient Reinforcement Finetuning for Large Toolspaces
- Reasoning and Tool-use Compete in Agentic RL:From Quantifying Interference to Disentangled Tuning
- Hammer: Robust Function-Calling for On-Device Language Models via Function Masking
- MCP-Universe: Benchmarking Large Language Models with Real-World Model Context Protocol Servers
- LiveMCPBench: Can Agents Navigate an Ocean of MCP Tools?
- gpt-oss-120b & gpt-oss-20b Model Card
- ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- ReAct: Synergizing Reasoning and Acting in Language Models
- Gradient Surgery for Multi-Task Learning
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems