|
|
|
Stories from November 17, 2023 (UTC)
|
| 1. |
Wikidata, with 12B facts, can ground LLMs to improve their factuality (arxiv.org) |
|
219 points by raybb on Nov 17, 2023 | hide | past | pdf | 84 comments
|
|
| 2. |
Can LLMs Follow Simple Rules? (arxiv.org) |
|
3 points by PaulHoule on Nov 17, 2023 | hide | past | pdf | 4 comments
|
|
| 3. |
Will AIs fake alignment during training in order to get power? (arxiv.org) |
|
2 points by wjb3 on Nov 17, 2023 | hide | past | pdf | discuss
|
|
| 4. |
Comparing Humans, GPT-4, and GPT-4V on Abstraction and Reasoning Tasks (arxiv.org) |
|
1 point by georgehill on Nov 17, 2023 | hide | past | pdf | discuss
|
|
| 5. |
Unprompted, LLM Agents can strategically deceive users when put under pressure (arxiv.org) |
|
1 point by famouswaffles on Nov 17, 2023 | hide | past | pdf | 1 comment
|
|
| 6. |
Performance Trade-Offs of Watermarking Large Language Models (arxiv.org) |
|
1 point by jruohonen on Nov 17, 2023 | hide | past | pdf | 1 comment
|
|
| 7. |
Contrastive Chain-of-Thought Prompting (arxiv.org) |
|
1 point by georgehill on Nov 17, 2023 | hide | past | pdf | discuss
|
|
|
|
