Vision-language navigation requires agents to follow natural-language instructions to reach target locations, making high-quality language supervision essential. Since human annotation is costly, prior work has explored scalable data synthesis methods. However, existing methods typically either use separate models for...
Song-Ze Li, Hao-Ming Gu, Zi-Yang Gong et al.· 0 citations
Latent communication passes internal states between language models instead of decoded text, but higher receiver accuracy does not show that the receiver used the message content, so the interface can supply the gain while making the sharer dispensable.
Lin-Quan Wu, Shi-Chang Meng, Tianxiang Jiang et al.· 1 citation
More privileged information does not always make a better teacher. We study this tension in on-policy self-distillation (OPSD), where a self-teacher scores the student's own rollouts under privileged context, conventionally a complete reference solution that bundles the final answer with one particular reasoning path....
Kang-Hui Tian, Si-Yuan Liu, Tianxiang Jiang et al.· 2 citations
Across seven benchmarks, TimeLens2-2B outperforms all size-matched baselines on every benchmark, while the 4B and 8B variants achieve state-of-the-art performance, surpassing open-source models with up to 397B parameters.
