readme updated
Browse files
README.md
CHANGED
|
@@ -8,11 +8,12 @@ It thinks like o1
|
|
| 8 |
[ ] Better error handling
|
| 9 |
[ ] Add Tools (web, math, code)
|
| 10 |
[ ] Make cli
|
| 11 |
-
|
| 12 |
|
| 13 |
## What it does
|
| 14 |
|
| 15 |
-
- It
|
|
|
|
| 16 |
|
| 17 |
## Installation
|
| 18 |
|
|
@@ -28,10 +29,31 @@ HAVE FUN.
|
|
| 28 |
|
| 29 |
## Helpful Papers
|
| 30 |
|
| 31 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 32 |
|
| 33 |
2. The Impact of Reasoning Step Length on Large Language Models
|
| 34 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
3. Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters [2212.10001](https://arxiv.org/abs/2212.10001)
|
| 36 |
```bibtex
|
| 37 |
@misc{wang2023understandingchainofthoughtpromptingempirical,
|
|
@@ -43,4 +65,8 @@ HAVE FUN.
|
|
| 43 |
primaryClass={cs.CL},
|
| 44 |
url={https://arxiv.org/abs/2212.10001},
|
| 45 |
}
|
| 46 |
-
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
[ ] Better error handling
|
| 9 |
[ ] Add Tools (web, math, code)
|
| 10 |
[ ] Make cli
|
| 11 |
+
[ ] better prompts for mathematical reasoning/reviewing
|
| 12 |
|
| 13 |
## What it does
|
| 14 |
|
| 15 |
+
- It taks the prompt, decides whether to use chain of thought or direct answer, if cot then generates answer and does self review, if direct answer then directly generates answer.
|
| 16 |
+
- Mathematical reasoning, symbolic reasoning and semi-symbolic reasoning kind of tasks generally improves with chain of thought, but direct answer is good for factual recall, simple inferences, commonsense reasoning, language understanding tasks.
|
| 17 |
|
| 18 |
## Installation
|
| 19 |
|
|
|
|
| 29 |
|
| 30 |
## Helpful Papers
|
| 31 |
|
| 32 |
+
1. To Cot or not to Cot? CHAIN-OF-THOUGHT HELPS MAINLY ON MATH AND SYMBOLIC REASONING
|
| 33 |
+
```bibtex
|
| 34 |
+
@misc{sprague2024cotcotchainofthoughthelps,
|
| 35 |
+
title={To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoning},
|
| 36 |
+
author={Zayne Sprague and Fangcong Yin and Juan Diego Rodriguez and Dongwei Jiang and Manya Wadhwa and Prasann Singhal and Xinyu Zhao and Xi Ye and Kyle Mahowald and Greg Durrett},
|
| 37 |
+
year={2024},
|
| 38 |
+
eprint={2409.12183},
|
| 39 |
+
archivePrefix={arXiv},
|
| 40 |
+
primaryClass={cs.CL},
|
| 41 |
+
url={https://arxiv.org/abs/2409.12183},
|
| 42 |
+
}
|
| 43 |
+
```
|
| 44 |
|
| 45 |
2. The Impact of Reasoning Step Length on Large Language Models
|
| 46 |
+
```bibtex
|
| 47 |
+
@misc{jin2024impactreasoningsteplength,
|
| 48 |
+
title={The Impact of Reasoning Step Length on Large Language Models},
|
| 49 |
+
author={Mingyu Jin and Qinkai Yu and Dong Shu and Haiyan Zhao and Wenyue Hua and Yanda Meng and Yongfeng Zhang and Mengnan Du},
|
| 50 |
+
year={2024},
|
| 51 |
+
eprint={2401.04925},
|
| 52 |
+
archivePrefix={arXiv},
|
| 53 |
+
primaryClass={cs.CL},
|
| 54 |
+
url={https://arxiv.org/abs/2401.04925},
|
| 55 |
+
}
|
| 56 |
+
```
|
| 57 |
3. Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters [2212.10001](https://arxiv.org/abs/2212.10001)
|
| 58 |
```bibtex
|
| 59 |
@misc{wang2023understandingchainofthoughtpromptingempirical,
|
|
|
|
| 65 |
primaryClass={cs.CL},
|
| 66 |
url={https://arxiv.org/abs/2212.10001},
|
| 67 |
}
|
| 68 |
+
```
|
| 69 |
+
|
| 70 |
+
# But me a Coffee
|
| 71 |
+
|
| 72 |
+
[](https://buymeacoffee.com/tikendraw)
|