Stronz

2 ms·

1. 1.▲
   Show HN: H-Core Snapshot – forcing LLMs to execute instead of explain(github.com)

   2 points by Stronz 6mo ago|0 comments
2. 2.▲
   by Stronz 6mo ago

   I built this after repeatedly running into the same issue in longer task-oriented conversations: the model gradually drifts into explanation mode and becomes verbose, speculative, and initiative-heavy. The idea is to initialize the conversa
3. 3.▲
   Show HN: H-Core Snapshot – prevents ChatGPT from drifting(github.com)

   1 points by Stronz 6mo ago|1 comments
4. 4.▲
   by Stronz 7mo ago

   One thing I noticed with local models is that conversation behavior tends to drift over time as well. Even when running locally, the model often starts structured but gradually becomes more verbose or explanatory in longer threads. Curious
5. 5.▲
   by Stronz 7mo ago

   It might also depend on how the tools are used. In practice a lot of value seems to come from reducing small bits of friction rather than dramatically increasing output.