Tweet fijado
Engram
15 posts
Engram
@EngramLab
scaling compute on your context
San Francisco Se unió Nisan 2026
1 Siguiendo3.9K Seguidores
Engram retuiteado
By scaling compute on user context, we reduce token spend. But it's about more than lowering cost. To develop expertise is to reduce the energy it takes to solve a problem, freeing capacity to solve harder problems yet
Was great chatting about this with the amazing @LM_Braswell!
Kleiner Perkins@kleinerperkins
The models we use every day are brilliant strangers. They forget your organization the moment a chat ends, then relearn it on the next query. @EngramLab fixes that. It learns your world once and reuses that memory, matching frontier systems on 1-10% of the tokens. @Microsoft, @NotionHQ, and @Harvey are already testing it within their organizations. Congratulations to the team, and hear directly from @dan_biderman (CEO and co-founder) and Sabri Eyuboglu (CTO and co-founder) with @LM_Braswell ⬇️
English
Dan and Sabri sat down with @LM_Braswell to talk about our technical direction. Thanks for having us!
Kleiner Perkins@kleinerperkins
The models we use every day are brilliant strangers. They forget your organization the moment a chat ends, then relearn it on the next query. @EngramLab fixes that. It learns your world once and reuses that memory, matching frontier systems on 1-10% of the tokens. @Microsoft, @NotionHQ, and @Harvey are already testing it within their organizations. Congratulations to the team, and hear directly from @dan_biderman (CEO and co-founder) and Sabri Eyuboglu (CTO and co-founder) with @LM_Braswell ⬇️
English
Engram retuiteado
The models we use every day are brilliant strangers. They forget your organization the moment a chat ends, then relearn it on the next query.
@EngramLab fixes that. It learns your world once and reuses that memory, matching frontier systems on 1-10% of the tokens.
@Microsoft, @NotionHQ, and @Harvey are already testing it within their organizations.
Congratulations to the team, and hear directly from @dan_biderman (CEO and co-founder) and Sabri Eyuboglu (CTO and co-founder) with @LM_Braswell ⬇️
Engram@EngramLab
English
Engram retuiteado
Kleiner Perkins@kleinerperkins
The models we use every day are brilliant strangers. They forget your organization the moment a chat ends, then relearn it on the next query. @EngramLab fixes that. It learns your world once and reuses that memory, matching frontier systems on 1-10% of the tokens. @Microsoft, @NotionHQ, and @Harvey are already testing it within their organizations. Congratulations to the team, and hear directly from @dan_biderman (CEO and co-founder) and Sabri Eyuboglu (CTO and co-founder) with @LM_Braswell ⬇️
English
Engram retuiteado
this is a wonderful group of people working on really interesting problems! me and @aslanpouthakoun had the very fun (and maybe a little intimidating) experience of giving a talk on finetuning at engram a few weeks ago, and we couldn’t have asked for a better audience
Engram@EngramLab
English
Engram retuiteado
Engram is one of the more sophisticated teams I’ve had the pleasure of working with at Modal
exciting launch, looking forward to assisting with even weirder deployments in the future!
Sabri Eyuboglu@EyubogluSabri
Modal's been super important for our velocity over the last 6 months - Training on each user's context means scaling out to thousands of GPUs in quick bursts. Modal allowed us to do this from day zero, before we could keep a large committed cluster hot - Our research team experiments with weird parameterizations all the time and needs to make changes to our inference and training servers. Modal makes it super easy for everyone on the team to deploy new endpoints for dogfooding and eval
English
Engram retuiteado
Engram retuiteado
Modal's been super important for our velocity over the last 6 months
- Training on each user's context means scaling out to thousands of GPUs in quick bursts. Modal allowed us to do this from day zero, before we could keep a large committed cluster hot
- Our research team experiments with weird parameterizations all the time and needs to make changes to our inference and training servers. Modal makes it super easy for everyone on the team to deploy new endpoints for dogfooding and eval
Akshat Bubna@akshat_b
It's been awesome getting to work with this team on their infra needs over the past few months. Congrats on the launch!
English
Having fun and liking your coworkers is such an underrated multiplier of research productivity.
Lots of stuff can objectively suck about building big models. Staying up late debugging and making sure the gpus are hot. Looking at the data. Having your fun ideas be a bust.
Having the trust of the people around you to be open the most charitable interpretation of your ideas who are in it with you is a big deal. It makes the shitty parts more bare able, and makes it easier to feel like everyone’s wins our the teams win.
English
clever name.
In neuroscience, an engram is the physical trace of a memory stored in the brain. The name is the thesis.
Engram@EngramLab
English
just to confirm @EngramLab - @eazyengram is not on the team??? congrats on the raise 👏
x.com/frontrunvc/sta…
Engram@EngramLab
English