Thinking Fast and Slow with Deep Learning and Tree Search · HackerPapers discussions

Thinking Fast and Slow with Deep Learning and Tree Search — mentioned by fabmilo

Sun, 10 Mar 2024 23:58:34 GMT

I believe this kind of graph exploration is what we need to progress reasoning in AI. Plain LLMS will fail. The link has tons of good references, including the Zobrist hashing https://en.wikipedia.org/wiki/Zobrist_hashing for game tables. We need to find a good hashing for language based state description so that graph exploration doesn't explode computationally. Another good read for Tree Search is Thinking Fast and Slow: https://arxiv.org/abs/1705.08439 and Teaching Large Language Models to Reason with Reinforcement Learning: https://arxiv.org/abs/2403.04642 comparing the MCTS approach to…

Paper and discussion history. A mention is not an endorsement.

Thinking Fast and Slow with Deep Learning and Tree Search — mentioned by nlperguiy

Thu, 28 Dec 2017 02:28:57 GMT

https://arxiv.org/abs/1705.08439 The original paper. The references in the paper paint a much clearer picture of where exactly the idea behind reinforcement learning with optimal, suboptimal, random oracles comes from. There are also mathematical proofs that these setups work. I was quite shocked to not see [6, 16] references in any of the recent MCTS papers. These references prove why the stuff works and show how well it works. But the whole field of imitation learning seems invisible to the deep RL papers. Don't have the faintest idea why. The algorithm described is the ultimate generalized…

Paper and discussion history. A mention is not an endorsement.

Thinking Fast and Slow with Deep Learning and Tree Search — mentioned by sanxiyn

Thu, 07 Dec 2017 13:41:53 GMT

Quick, someone apply AlphaZero algorithm to this MCTS prover for math: https://arxiv.org/abs/1611.05990 I actually prefer "expert iteration" and "tree-policy target" term instead of "AlphaZero algorithm", because they emphasize what is novel about it among the entire system. Terms are from https://arxiv.org/abs/1705.08439

Paper and discussion history. A mention is not an endorsement.