Open-source reflection lama 70B beats Claude 3.5 and GPT-4 on benchmarks (reflectionllama.com) 6 points by oldcai 1y ago ↗ HN
[–] magicalhippo 1y ago ↗ Related recent discussion:Reflection 70B, the top open-source modelhttps://news.ycombinator.com/item?id=41459781 [–] oldcai 1y ago ↗ That's a great conversation about the reflection 70b, but do you still have doubts about whether it's a hype or a game-changer?The link provided leads to a playground for the reflection llama 70B. [–] magicalhippo 1y ago ↗ In the Physics of Language Models talk[1], he shows how a LLM trained to be able to backtrack can give much better answers.However he also points out it has to be included in the initial training, you can't improve a non-backtrack-trained model by finetuning it later.So seems it's probably the way to go for training new models, but limited applicability to those already trained.[1]: https://www.youtube.com/watch?v=yBL7J0kgldU
[–] oldcai 1y ago ↗ That's a great conversation about the reflection 70b, but do you still have doubts about whether it's a hype or a game-changer?The link provided leads to a playground for the reflection llama 70B. [–] magicalhippo 1y ago ↗ In the Physics of Language Models talk[1], he shows how a LLM trained to be able to backtrack can give much better answers.However he also points out it has to be included in the initial training, you can't improve a non-backtrack-trained model by finetuning it later.So seems it's probably the way to go for training new models, but limited applicability to those already trained.[1]: https://www.youtube.com/watch?v=yBL7J0kgldU
[–] magicalhippo 1y ago ↗ In the Physics of Language Models talk[1], he shows how a LLM trained to be able to backtrack can give much better answers.However he also points out it has to be included in the initial training, you can't improve a non-backtrack-trained model by finetuning it later.So seems it's probably the way to go for training new models, but limited applicability to those already trained.[1]: https://www.youtube.com/watch?v=yBL7J0kgldU
[–] morbicer 1y ago ↗ I must be missing something but the model is absolutely useless because it's giving unrelated answers.Check the questions people posted:https://reflectionllama.com/discover
5 comments
[ 2.0 ms ] story [ 26.7 ms ] threadReflection 70B, the top open-source model
https://news.ycombinator.com/item?id=41459781
The link provided leads to a playground for the reflection llama 70B.
However he also points out it has to be included in the initial training, you can't improve a non-backtrack-trained model by finetuning it later.
So seems it's probably the way to go for training new models, but limited applicability to those already trained.
[1]: https://www.youtube.com/watch?v=yBL7J0kgldU
Check the questions people posted:
https://reflectionllama.com/discover