FrancescoMassa

6 ms·

1.
The Rise of HOMs: Harmful Open Models(twitter.com)
2 points by FrancescoMassa 14d ago|0 comments
2.
The Rise of HOMs: Harmful Open Models(x.com)
4 points by FrancescoMassa 15d ago|0 comments
3.
Is someone Into process mining? Let's talk in this thread
1 points by FrancescoMassa 2mo ago|0 comments
4.
by FrancescoMassa 2mo ago
it works exactly like that...that's why you're seeing the notice con claude code when switching models
5.
by FrancescoMassa 2mo ago
You transfer context from the kv cache of model A to the kv cache of model B
6.
by FrancescoMassa 2mo ago
Actually yes because context is kept on kV cache
7.
by FrancescoMassa 2mo ago
For sure that’s absolutely true because you’re switching context from a kV cache to another one. We introduced 2 algorythms for solving this : 1) sticky : switches model only when convenient 2) smartsqueeze : fast advanced compression for r
8.
How we're saving 50k$ of Claude API every month(github.com)
3 points by FrancescoMassa 2mo ago|9 comments
9.
by FrancescoMassa 2mo ago
Brick is a cheap & smart spatial router that plugged into claude code redirects the query to the best model in your pool. You can use it to save tokens, use other providers in claude code without changing models by hand or making your p
10.
Claude Savings with context caching awareness(github.com)
2 points by FrancescoMassa 2mo ago|1 comments
11.
by FrancescoMassa 2mo ago
We have a new update for brick : with feedbacks from users we inserted context caching aware routing. so that savings are improved even if switching models transfers context from one kv cache to another. we introduced a tier1 compaction alg
12.
by FrancescoMassa 3mo ago
On our tests subagents & well used workflows are 20-30% more expensive for context & token efficiency
13.
Save Claude Code Tokens with Smart Routing(github.com)
11 points by FrancescoMassa 3mo ago|3 comments
14.
Brick: SOTA LLM Routing(arxiv.org)
3 points by FrancescoMassa 3mo ago|0 comments