Excited to be judging the biggest hackathon at #SFTechWeek with @MongoDB and fellow speakers from @aiatMeta
, @WorkOS
, 1Password, Coinbase, @FireworksAI_HQ et al! https://
fandf.co/42v5oS2 One of the biggest topics this year is Memory and Context Engineering for Agents. Over 1
LLMS
-

Judging SF Tech Week Hackathon with Major Tech Partners
By
–
-
Build Autograd Engine, Mini-GPT, and LoRA Fine-tuning From Scratch
By
–
– build an autograd engine from scratch
– write a mini-GPT from scratch
– implement LoRA and fine-tune a model on real data
– hate CUDA at least once
– cry
– keep going the roadmap – 5 phases
– if you already know something? skip
– if you're lost? rewatch
– if you’re stuck? use -
GLM-4.5 Air Local Deployment for Sensitive Data Projects
By
–
yeah i know, i am using glm-4.5 air locally for some sensitive data project otherwise that plan is great for its price they're also dropping 4.6 air soon
-

DeepThink IMO Gold Model Achieves State-of-the-Art FrontierMath Results
By
–
Very cool to see this new state-of-the-art result on FrontierMath achieved by the #DeepThink IMO Gold model that we built 3 months (a long time) ago 🙂 The fact that is evaluated externally using a publicly available API is a strong testament that what we built generalizes beyond
-
AI Models Moving to Devices: Key Considerations
By
–
ai models might soon live in your device, not just in giant datacenters. if you’re a founder / engineer / prompt architect, start thinking:
→ designing models aware of analog constraints
→ hybrid architectures: analog + digital fallbacks
→ compression, pruning, or -
Offline AI possible with analog in-memory computing
By
–
HOLY SH*T. OFFLINE AI IS POSSIBLE? Researchers published a new analog in-memory computing (IMC) architecture in Nature Computational Science, targeting the transformer attention mechanism. Nature Instead of shuffling data between memory and compute (the classic GPU bottleneck),
-
Claude Code Implements Plugin Pattern for AI Integration
By
–
And today it's Claude Code's implementation of a similar pattern, which they're calling plugins https://
x.com/claudeai/statu
/claudeai/status/1976332881409737124
… -
Gemini Extensions Release Highlights Rapid Coding Agent Tool Competition
By
–
It's interesting how close together the feature releases are from the different coding agent tools – here's Gemini extensions from yesterday https://
x.com/googlecloudtec
/googlecloudtech/status/1975931556804935870
… -
Open weights models lag behind proprietary AI advancement
By
–
feels like ages since a new truly paradigm shifting model released on open weights side? or am I missing something? the closest on proprietary side is gpt-5-codex
-
RTX 5090 vs 4x 3090s VRAM comparison for LLM inference
By
–
5090 has 32GB of VRAM, 4x 3090s have 96GB of VRAM when it comes to LLMs inference, we care more about memory as models are better fully offloaded into VRAM than being shared across system RAM and a single RTX 5090's VRAM