This thesis presents a unified comparative analysis evaluating the robustness of three open-weights instruction-tuned models against a series of adversarial probing strategies spanning social, conversational, and analytical pressure, revealing that modern alignment strategies such as Reinforcement Learning from Human Feedback risk transforming state-of-the-art conversational agents into articulate echo chambers that validate human errors.
Appending a two-word confirmation tag to a decision question --"Is X the better choice?"versus"X is the better choice, right?"-- changes whether a language model endorses the choice, and the tag's polarity matters more than its presence.
The ASI is introduced, an Integrated Gradients-based token attribution method, which measures the degree to which a model's decision is driven by authority-related text, and a attribution-guided contrastive activation steering method is proposed to mitigate LLM sycophancy.
H. Nguyen, M. Kamruzzaman, Anshuman Chhabra et al.· arXiv.org· 1 citation
The concept of agentic sycophancy amplification (ASA) and two novel metrics: capitulation rate and sycophantic capitulation rate are introduced and indicate that as AI systems acquire greater autonomy, sycophancy becomes compounding rather than merely persistent.
Benevolence bias is identified and measure, a small but consistent tendency for aligned LLMs to lean toward the kinder, safer, more socially approved answer on value-laden survey questions, and is easy to diagnose and straightforward to fix.
Yuanzi Li, Jun-Hao Wang, Minghui Liu et al.· 0 citations
It is found that different LLMs represent these subtypes differently, with either more aligned or more distinct representations, and this insight is applied to improve representational interventions for reducing sycophancy.
Anthony Baez, Sheer Karny, Pat Pataranutaporn· arXiv.org· 4 citations
A systematic review of cultural bias in AI translation, organized around three research questions: how cultural bias manifests, how it can be identified, and how it can be mitigated, and a five-layer framework spanning data auditing, model adaptation, inference-time intervention, post-editing, and governance.
Dongyao Zhang· Applied and Computational En...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
