penn strategic projects @afterquery
San Francisco, CA
Joined September 2023
- Tweets300
- Following250
- Followers610
- Likes884
Lovely work by the AQ Team
@AfterQuery is partnering with @NVIDIA to help enterprises post-train Nemotron 3.5 Super VL.
The future of enterprise agents is open. 馃挌
Anderson Beck retweeted
@AfterQuery is partnering with @NVIDIA to help enterprises post-train Nemotron 3.5 Super VL.
The future of enterprise agents is open. 馃挌
Enterprise models.
Our research team post-trained an open model with @hebbia to approach Astra performance in finance, at 1/7th the cost.
Enterprise-specific frontier intelligence is becoming increasingly affordable.
Anderson Beck retweeted
Our research team post-trained an open model with @hebbia to approach Astra performance in finance, at 1/7th the cost.
Enterprise-specific frontier intelligence is becoming increasingly affordable.
@AfterQuery just moved an open model past the cost-accuracy frontier in finance.
We post-trained @NVIDIAAI Nemotron 3.5 Super VL on expert credit tasks with @hebbia.
Perfect solved: 47% to 74%. That beats Sol 6, and lands within 3 points of Astra 6, at about 1/7 the cost.
Link to the full case study in replies.
Enterprise-specific models are the future
Every serious enterprise will soon own and control its own intelligence.
Huge congratulations to the @datadoghq team on Scout.
Anderson Beck retweeted
Every serious enterprise will soon own and control its own intelligence.
Huge congratulations to the @datadoghq team on Scout.
Meet Scout 1.0, our first AI model built for investigating production incidents.
Scout matches frontier-model investigation quality at a fraction of the cost, which makes it practical to investigate more issues, earlier.
We trained Scout on realistic simulations built from real-world incidents and production telemetry, using reinforcement learning to reward accurate root causes backed by evidence.
Learn more and sign up for the preview: bit.ly/3TJknGO
Anderson Beck retweeted
Proud of the team pushing what open models can do in finance.
@AfterQuery just moved an open model past the cost-accuracy frontier in finance.
We post-trained @NVIDIAAI Nemotron 3.5 Super VL on expert credit tasks with @hebbia.
Perfect solved: 47% to 74%. That beats Sol 6, and lands within 3 points of Astra 6, at about 1/7 the cost.
Link to the full case study in replies.
Amazing work by the team on this!
@AfterQuery just moved an open model past the cost-accuracy frontier in finance.
We post-trained @NVIDIAAI Nemotron 3.5 Super VL on expert credit tasks with @hebbia.
Perfect solved: 47% to 74%. That beats Sol 6, and lands within 3 points of Astra 6, at about 1/7 the cost.
Link to the full case study in replies.
Anderson Beck retweeted
@AfterQuery just moved an open model past the cost-accuracy frontier in finance.
We post-trained @NVIDIAAI Nemotron 3.5 Super VL on expert credit tasks with @hebbia.
Perfect solved: 47% to 74%. That beats Sol 6, and lands within 3 points of Astra 6, at about 1/7 the cost.
Link to the full case study in replies.
be there
Anderson Beck retweeted
@AfterQuery鈥檚 throwing a sftw happy hour with @NVIDIAAI tonight at 6:30pm.
reply for an invite.
beautiful research
One of the best decisions we鈥檝e made early at @AfterQuery was building an in-house research team.
Knowing what great data looks like is table stakes. Raising the bar means showing labs their model鈥檚 losses, then providing data to close those gaps.
Anderson Beck retweeted
One of the best decisions we鈥檝e made early at @AfterQuery was building an in-house research team.
Knowing what great data looks like is table stakes. Raising the bar means showing labs their model鈥檚 losses, then providing data to close those gaps.
Not all coding data is created equal.
Using only 500 tasks from @AfterQuery's SWE agent dataset, our internal post-training team was able to improve Qwen3.8-27B-Medium by 11.3 points in only 15 steps of GRPO.
Read more about our training experiment in the blog post linked in the comments.
better data --> better models
every time
Not all coding data is created equal.
Using only 500 tasks from @AfterQuery's SWE agent dataset, our internal post-training team was able to improve Qwen3.8-27B-Medium by 11.3 points in only 15 steps of GRPO.
Read more about our training experiment in the blog post linked in the comments.
Anderson Beck retweeted
Not all coding data is created equal.
Using only 500 tasks from @AfterQuery's SWE agent dataset, our internal post-training team was able to improve Qwen3.8-27B-Medium by 11.3 points in only 15 steps of GRPO.
Read more about our training experiment in the blog post linked in the comments.
Anderson Beck retweeted
Progress is moving much faster than most people appreciate. So incredible to watch this happen in real time.
We鈥檙e releasing a broad range of new mathematical results produced by an internal frontier model.
We鈥檝e been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results.
github.com/openai/math
