# Edge Cases by AGI, Inc. #004 with Coleman Hooper @ UC Berkeley (BAIR) · Thu, Nov 19, 2026 · San Francisco · Free

Source: https://events.trymimetic.com/events/sf/edge-cases-by-agi-inc-004-with-coleman-hooper-uc-berkeley-bair-nov-2026

High Signal Event

# Edge Cases by AGI, Inc. #004 with Coleman Hooper @ UC Berkeley (BAIR)

📅
November 19, 2026

🕐
5:30 PM PT

📍
San Francisco

🎟️
Free · Approval required

## What we're seeing

- **Approval required · venue shared on acceptance**

## About This Event

Edge Cases is AGI, Inc.'s invite-only research night on what it takes to run real intelligence on-device.

**Speaker**

**[Coleman Hooper](https://chooper1.github.io/)**, postdoctoral scholar at UC Berkeley (BAIR and SLICE Lab). PhD at Berkeley, advised by Kurt Keutzer and Sophia Shao. Works on AI systems, efficient LLM inference and hardware-software co-design for ML. [LinkedIn](https://www.linkedin.com/in/coleman-hooper-165061193/)

**Talk**

**Efficient edge inference through parallel and looped generation**

In the cloud, LLM inference batches across many users. On a device there is one user, so decoding runs at batch size one and most of the hardware's parallel compute goes unused. Coleman will cover two ways around this. Diffusion language models refine several tokens in parallel, and his work trains them to need fewer refinement steps. Looped transformers pass each token through the model multiple times, and self-speculative decoding uses the earlier loops for drafting and the later loops for verification. The talk ends with recent work that combines the two for speedups above 10x.

Papers: [MLSys 2026](https://proceedings.mlsys.org/paper_files/paper/2026/hash/054de805fcceb78a201f5e9d53c85908-Abstract-Conference.html) · [Springer chapter](https://link.springer.com/chapter/10.1007/978-3-031-85747-8_2) · one more on arXiv by the talk

**Details**

- Thursday, November 19, 5:30 PM
- Our founders' house in San Francisco. Exact address shared on approval.
- 30 to 40 minute talk, off-the-record Q&A, then dinner with the group
- For AI researchers, PhD students and research engineers
- Seats are limited and registration is approval-based

More about the series: [theagi.company/edge-cases](https://theagi.company/edge-cases)

Want to present your on-device research at a future Edge Cases? [Apply here](https://form.typeform.com/to/AZDHOWgq?utm_source=luma)

## Hosted by

[AGI, Inc.](https://luma.com/agi-inc)

[Juliana C](https://luma.com/user/agijuliana)

[C K](https://luma.com/user/khotchaitanya)

[Div Garg](https://luma.com/user/usr-V8Q9em5gLS6D3Tt)

[Steve Frey](https://luma.com/user/usr-GZmY3BsrZPXbOrF)

[Jordan Ruppert](https://luma.com/user/usr-c2u1VJEnKHdzdpC)

[Naman Garg](https://luma.com/user/usr-L4aaWFK9T8GZUFg)

[Adam Awale](https://luma.com/user/usr-0AoMIIgeQWYfgV5)

[AGI Inc.](https://luma.com/user/agi_inc)

## Location

📍 San Francisco, CA

AI

Networking

Founders

Talk

GenAI

Research

[Register for This Event](https://lu.ma/edge-cases-004?utm_source=trymimetic&utm_medium=calendar&utm_campaign=ai_events)

[99 more SF AI events this week](/events/ai/sf)

Free Growth Analysis

## Get a free growth analysis for your company

See how your website, messaging, and go-to-market strategy stack up, in minutes.

[Get My Free Analysis](/?utm_source=events&utm_medium=referral&utm_campaign=ai&utm_content=event_detail_cta)

## More SF events like this

### [AIE x MLH Hackathon @ Code Summit](/events/sf/aie-x-mlh-hackathon-code-summit-nov-2026)

Today · 9:00 AM PT
San Francisco, CA
102 attending

### [Do Not Disturb HACKATHON](/events/sf/do-not-disturb-hackathon-nov-2026)

Today · 9:30 AM PT
San Francisco, CA

### [The Multimodal Hackathon](/events/sf/the-multimodal-hackathon-nov-2026)

Today · 2:00 PM PT
FOCUS Innovation Studio, 181 2nd St, San Francisco, CA 94105, USA

High Signal

### [SGLang Summit Workshop by Modal: Inference from First Principles](/events/sf/sglang-summit-workshop-by-modal-inference-from-first-principles-nov-2026)

Today · 2:50 PM PT
Fort Mason, San Francisco, CA 94109, USA
90 attending

[← Back to San Francisco Events](/events/ai/sf)
