Chengshuai Zhao
@chengshuaizhao.bsky.social
Pinned
📢 Excited to share our new arXiv preprint: "Is Chain-of-Thought Reasoning of LLMs a Mirage? A Data Distribution Lens"
📄 Read the full paper: arxiv.org/abs/2508.01191
💻 Code & data: github.com/ChengshuaiZh...
🤗 Huggingface: huggingface.co/papers/2508....
📄 Read the full paper: arxiv.org/abs/2508.01191
💻 Code & data: github.com/ChengshuaiZh...
🤗 Huggingface: huggingface.co/papers/2508....
📢 Excited to share our new arXiv preprint: "Is Chain-of-Thought Reasoning of LLMs a Mirage? A Data Distribution Lens"
📄 Read the full paper: arxiv.org/abs/2508.01191
💻 Code & data: github.com/ChengshuaiZh...
🤗 Huggingface: huggingface.co/papers/2508....
📄 Read the full paper: arxiv.org/abs/2508.01191
💻 Code & data: github.com/ChengshuaiZh...
🤗 Huggingface: huggingface.co/papers/2508....
September 7, 2025 at 6:01 AM
📢 Excited to share our new arXiv preprint: "Is Chain-of-Thought Reasoning of LLMs a Mirage? A Data Distribution Lens"
📄 Read the full paper: arxiv.org/abs/2508.01191
💻 Code & data: github.com/ChengshuaiZh...
🤗 Huggingface: huggingface.co/papers/2508....
📄 Read the full paper: arxiv.org/abs/2508.01191
💻 Code & data: github.com/ChengshuaiZh...
🤗 Huggingface: huggingface.co/papers/2508....
Reposted by Chengshuai Zhao
Is Chain-of-Thought Reasoning of LLMs a Mirage? A Data Distribution Lens
https://arstechnica.com/ai/2025/08/researchers-find-llms-are-bad-at-logical-inference-good-at-fluent-nonsense/
https://news.ycombinator.com/item?id=44872850
https://arstechnica.com/ai/2025/08/researchers-find-llms-are-bad-at-logical-inference-good-at-fluent-nonsense/
https://news.ycombinator.com/item?id=44872850
LLMs’ “simulated reasoning” abilities are a “brittle mirage,” researchers find
Chain-of-thought AI “degrades significantly” when asked to generalize beyond training.
arstechnica.com
August 12, 2025 at 1:30 PM
Is Chain-of-Thought Reasoning of LLMs a Mirage? A Data Distribution Lens
https://arstechnica.com/ai/2025/08/researchers-find-llms-are-bad-at-logical-inference-good-at-fluent-nonsense/
https://news.ycombinator.com/item?id=44872850
https://arstechnica.com/ai/2025/08/researchers-find-llms-are-bad-at-logical-inference-good-at-fluent-nonsense/
https://news.ycombinator.com/item?id=44872850