Search papers, labs, and topics across Lattice.
3
0
6
3
This work introduces HarnessVLN, a zero-shot, training-free framework that unifies instruction-following and object-goal navigation through a shared Agent Harness, and demonstrates its applicability to both navigation tasks in real-world environments.
Faulty-code-driven test synthesis boosts code generation performance by 3% in LLMs, tackling reward hacking and validation issues head-on.
Most coding agents fail to proactively fix bugs without issue reports, revealing a critical gap in their capabilities.
