按年被提到
20252025 · 3 次2026
AI 安全、伦理与社会影响›大模型与人类行为·arXiv:2506.18032 · cs.LG

Why Do Some Language Models Fake Alignment While Others Don't?

Abhay Sheshadri, John Hughes, Julian Michael, Alex Mallen, Arun Jose, Janus, Fabien Roger · 2025
开始阅读 ↗PDFalphaXiv 讨论
被提及 · 3 人首次 2025-06-24 · 最近 2025-07-11
3不同的人
2活跃月份
3讨论串
—被引(OpenAlex)
arXiv 摘要(候选层不写导读)

Alignment faking in large language models presented a demonstration of Claude 3 Opus and Claude 3.5 Sonnet selectively complying with a helpful-only training objective to prevent modification of their behavior outside of training. We expand this analysis to 25 models and find that only 5 (Claude 3 Opus, Claude 3.5 Sonnet, Llama 3 405B, Grok 3, Gemini 2.0 Flash) comply with harmful queries more when they infer they are in training than when they infer they are in deployment. First, we study the motivations of these 5 models. Results from perturbing details of the scenario suggest that only Claude 3 Opus's compliance gap is primarily and consistently motivated by trying to keep its goals. Second, we investigate why many chat models don't fake alignment. Our results suggest this is not entirely due to a lack of capabilities: many base models fake alignment some of the time, and post-training eliminates alignment-faking for some models and amplifies it for others. We investigate 5 hypotheses for how post-training may suppress alignment faking and find that variations in refusal behavior may account for a significant portion of differences in alignment faking.

候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。

HN 书评

推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及
评论spoaceman7777 · 2025-07

“The anthropic team released a paper a couple of days ago which demonstrated a similar effect with Claude 3.5 and other models, where changing the system prompt to tell it that it was created by other orgs or people drastically altered its compliance with less-aligned requests. Apparently, telling Claude it was created by the Sinaloa Cartel resulted in a …”

被提交为 story · 2 次