Large Language Models (LLMs) are increasingly used in embodied AI for robotic manipulation and planning, but their deployment in physical systems exposes vulnerabilities such as jailbreak attacks that induce unsafe behavior. Existing inference-time defenses are mainly evaluated on short instructions containing a single...
Ali Taheri, Alireza Taban, Qi-Zhou Wang et al.· 0 citations
Longer contexts allow large language models (LLMs) to condition their predictions on more information, but inference costs grow prohibitively with context length. An emerging alternative uses test-time updates to encode context as temporary changes to model weights. However, encoding this information into weights that...
Yue Wang, Qi-Zhou Wang, Feng Liu et al.· 0 citations
Tool-using large language model (LLM) agents turn generated text into real side effects, so poisoned tool metadata, retrieved pages, memory, and reusable skills can steer the next call. Vetting an artifact before admission does not settle this. A safe variant and a leaking variant can produce the same admission evidenc...
Feng-Peng Li, Qi-Zhou Wang, Yu-Ke Hu et al.· 1 citation
This analysis gives sufficient conditions under which suppression at supervised single-turn contexts yields a bound on multi-turn trajectory risk, and characterizes contraction relative to a base-policy risk budget evaluated on the trained policy's contexts.
Feng-Peng Li, Ke-Mou Li, Qi-Zhou Wang et al.· 0 citations
Pre-training equips large language models (LLMs) with a broad repertoire of behavioral patterns associated with roles, styles, values, and goals. Post-training teaches conditional enactment and makes a helpful Assistant the default, but it does not erase alternative modes from the weights; explicit prompts can therefor...
Ke-Mou Li, Zhuan Shi, Qi-Zhou Wang et al.· 3 citations
Whether unlearning runs exhibit exploitable structure in weight space is investigated, and it is observed that models from different runs still lie in a shared evaluation-performance basin, suggesting that stronger models may be recovered through an unlearning-tailored soup strategy, reducing the need for repeated tuni...
A new DG framework is introduced that focuses on learning partially shared features (PSFs)-features shared among subsets of source domains, which contain and generalize ESFs, which contain and generalize ESFs.
Zeng-Mao Wang, Qi-Zhou Wang, Chao-Yang Zhou et al.· IEEE Transactions on Pattern...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.