Google AI Introduces Consistency Training for Safer Language Models Under Sycophantic and Jailbreak Style Prompts
"Language models" 相关文章
最具争议性研究：大模型中间层输出可 100% 反推原始输入
3万字长文！通俗解析大语言模型LLM原理
Why do AI models use so many em-dashes?
自变量机器人王潜：具身智能 Scaling law 还有多远？ | 云启实干派
How Do We Evaluate the Quality of LLMs' Mathematical Responses?
When Will AI Transform the Economy?
[CS 2881r] Can We Prompt Our Way to Safety? Comparing System Prompt Styles and Post-Training Effects on Safety Benchmarks
Q2 AI Benchmark Results: Pros Maintain Clear Lead
RL记得更牢，SFT更健忘？普林斯顿陈丹琦团队改写后训练认知
Reducing the Probability of Undesirable Outputs in Language Models Using Probabilistic Inference
VLA/世界模型/WA/端到端是宣传分歧, 不是技术路线分歧
数字化迎来能动性拐点：创业者有哪些可以把握的系统性机会？｜奇绩大模型笔记 #8
Learning to Interpret Weight Differences in Language Models
Expanding Model Choice in VS Code with Bring Your Own Key
本日最好笑研究： 在训练 AI 的时候，如果强迫 AI 大量阅读社交媒体，会对 AI 造成不可逆的脑损伤。 与在高质量数据上训练的对照组相比，持续投喂垃圾数据（确切...
一致性轨迹强化学习登场：上海AI Lab让扩散语言模型实现全并行少步数推理
Catch Your Breath: Adaptive Computation for Self-Paced Sequence Production
Catch Your Breath: Adaptive Computation for Self-Paced Sequence Production
Catch Your Breath: Adaptive Computation for Self-Paced Sequence Production
