yogthos

https://archive.ph/50Q2X

https://archive.ph/50Q2X

Instead of just generating the next response, it simulates entire conversation trees to find paths that achieve long-term goals.

How it works:

• Generates multiple response candidates at each conversation state
• Simulates how conversations might unfold down each branch (using the LLM to predict user responses)
• Scores each trajectory on metrics like empathy, goal achievement, coherence
• Uses MCTS with UCB1 to efficiently explore the most promising paths
• Selects the response that leads to the best expected outcome

Limitations:

• Scoring is done by the same LLM that generates responses
• Branch pruning is naive - just threshold-based instead of something smarter like progressive widening
• Memory usage grows with tree size, there currently no node recycling

Instead of just generating the next response, it simulates entire conversation trees to find paths that achieve long-term goals.

How it works:

• Generates multiple response candidates at each conversation state
• Simulates how conversations might unfold down each branch (using the LLM to predict user responses)
• Scores each trajectory on metrics like empathy, goal achievement, coherence
• Uses MCTS with UCB1 to efficiently explore the most promising paths
• Selects the response that leads to the best expected outcome

Limitations:

• Scoring is done by the same LLM that generates responses
• Branch pruning is naive - just threshold-based instead of something smarter like progressive widening
• Memory usage grows with tree size, there currently no node recycling

https://www.garda.com/en-ca