ClawGym II: Exploring Black-Box RL on Agent Harness
Agent harnesses have substantially improved performance on long-horizon tasks by coordinating agent interactions with the environment. However, reinforcement learning through complex harnesses remains largely unexplored, as scaling such training to long-horizon agent tasks introduces fundamental challenges. In this work, we present a unified black-box RL framework for stable and scalable optimization of general agents through complex harnesses. Concretely, we first build a sandbox-based execution infrastructure that isolates task environments and harnesses within temporary sandboxes for large-
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- FuzzySimilar title/name (fuzzy) · 59%AgentCore-8B →
“Fuzzy title match (0.73): “ClawGym II: Exploring Black-Box RL on Agent Harness” ≈ “AgentCore-8B””
- PossiblePossibly related (embedding) · 26%camel-ai/owl →
“Possibly related via embedding similarity 0.56 (not asserted). Timestamp check: artifact slightly before paper (-27d).”
- PossiblePossibly related (embedding) · 60%Why sandboxing your agent is not enough →
- FuzzySimilar title/name (fuzzy) · 87%SWE-agent/SWE-agent →
“Fuzzy title match (0.94): “ClawGym II: Exploring Black-Box RL on Agent Harness” ≈ “SWE-agent/SWE-agent””
- FuzzySimilar title/name (fuzzy) · 87%zhayujie/CowAgent →
“Fuzzy title match (0.94): “ClawGym II: Exploring Black-Box RL on Agent Harness” ≈ “zhayujie/CowAgent””
- FuzzySimilar title/name (fuzzy) · 87%strands-agents/harness-sdk →
“Fuzzy title match (0.94): “ClawGym II: Exploring Black-Box RL on Agent Harness” ≈ “strands-agents/harness-sdk””
- FuzzyOverlapping authors or contributors · 62%DietrichGebert/ponytail →
“Shared author/contributor keys: cheng”
- FuzzySimilar title/name (fuzzy) · 59%NousResearch/hermes-agent →
“Fuzzy title match (0.73): “ClawGym II: Exploring Black-Box RL on Agent Harness” ≈ “NousResearch/hermes-agent””