
Member of Technical Staff @AnthropicAI 🤖 @StanfordEng @uwcse 👨🎓 neovim enthusiast 💻 I help neural networks find local minima 🧠
San Francisco, CA
Joined April 2009
- Tweets223
- Following378
- Followers559
- Likes2.5K
Dillon Laird retweeted
We asked an unreleased research version of Claude to take a stab at the Riemann hypothesis.
It didn’t solve it, but it did make strides on a related problem: it increased the lower bound for the fraction of zeros of the Riemann zeta function that satisfy the hypothesis from 41.6% to 67.2%.
anthropic.com/research/riema…
Dillon Laird retweeted
This is a super exciting release - Claude Fable 5 is the same underlying model as Mythos but with added safeguards. The benchmarks are great and it's SOTA on everything by a margin but I'll add that *qualitatively* also, this is a major-version-bump-deserving step change forward (imo of the same order as Claude 4.5 was in November), peaking especially for long problem-solving sessions on very difficult problems. You can give it a lot more ambitious tasks than what you're used to, the model "gets it" and it will just go, and it's never felt this tempting to stop looking at the code at all (but don't do this in prod!). The model still has quirks that people will run into and the safeguards are configured to be a little too trigger happy for launch, which can hopefully be tuned over time.
I feel a lot of things changing as working software increasingly comes out on a tap. The Jevon's paradox kicks in and I feel my own demand for software growing substantially. You can ask for anything - explainers, visualizers, dashboards, bespoke single-use apps (e.g. a full wandb that is hyper-specific just for your project), you can 10X your test suite, auto-optimize code, run giant research projects with custom HTML for the results, anything! "Free your mind" (Matrix ref). Really looking forward to all the things people build!
Replying to @claudeai
Fable 5 is state-of-the-art on nearly all tested benchmarks, with exceptional performance in software engineering, knowledge work, scientific research, and vision.
The longer and more complex the task, the larger Fable 5’s lead over our other models.
Dillon Laird retweeted
over the weekend i had another obvious thing to check, namely whether claude autonomously resolves the famed sum-product conjecture over the reals. answer: yes
Introducing Claude Opus 4.8: it builds on Opus 4.7 with sharper judgment, more honesty about its own progress, and the ability to work independently for longer than its predecessors.
Available today at the same price.
Dillon Laird retweeted
Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.
Dillon Laird retweeted
I always thought the decline in fundamental AI research funding would happen because AI didn’t generate enough value to be worth the cost.
But it seems like it’s happening because it generated too much value. And the race to capture that value is taking priority.
Just remembering that a lot of this started in curiosity driven industry research labs.
Dillon Laird retweeted
A decision on SB-1047 is due soon. Governor @GavinNewsom has said he's concerned about its "chilling effect, particularly in the open source community". He's right, and I hope he will veto this.
If you agree, please like/retweet this to show your support for VETOing SB-1047!
Dillon Laird retweeted
Introducing Meta Segment Anything Model 2 (SAM 2) — the first unified model for real-time, promptable object segmentation in images & videos.
SAM 2 is available today under Apache 2.0 so that anyone can use it to build their own experiences
Details ➡️ go.fb.me/p749s5
Dillon Laird retweeted
Vision-language models benefit from hyperbolic embeddings for standard tasks, but did you know that hyperbolic vision-language models also have surprising properties?
Our new #TMLR paper shows 3 intriguing properties.
w/ @sarahibrahimi_ @GhadimiAtigMina @nannevn @marcelworring
Dillon Laird retweeted
Forget any document retrievers, use ColPali 💥💥
Document retrieval is done through OCR + layout detection, but it's overkill and doesn't work well! 🤓
ColPali uses a vision language model, which is better in doc understanding 📑
keep reading ⤋
Dillon Laird retweeted
Dive into the future with large multimodal models! These AI marvels go beyond text, bringing technology closer to understanding the world as we do. It's not just innovation; it's a revolution. Explore the possibilities:
youtube.com/watch?v=UXNvodLq…
#AIRevolution #MultimodalAI
Dillon Laird retweeted
After designing the imaging acquisition and data labeling, here's my third blog that talks about model training and evaluation. Based on our past experience at Landing AI we have developed some best practices that are aligned with our #DataCentric and #MLOps method.
Before deploying a model into production, #MachineLearning engineers will need to implement model training and evaluation. @ivanzhouyq shares best practices in our new blog post: https://xitter.gq/t.co/hxWZkZYlpj #data
Dillon Laird retweeted
I've always favored workplace transparency & professional well being. With partnership from @hashimovrustam, our passion project to help this cause is real & live!
Join us & help bring about a satisfied workforce & transparent workplaces with app.everi.one
#transparency
Dillon Laird retweeted
6 keys to creating quality images needed for training your model:
✅ Lighting
✅ Background
✅ Field of View
✅ Exposure
✅ Depth of Field
✅ Color vs black and white
Learn more: https://xitter.gq/t.co/0rGzUAF7Fd @DillonLaird @ivan_thinks #MachineVision #ComputerVision #AI
Dillon Laird retweeted
Developing an AI application for real-world problems is challenging: you have to optimize not only ML models, but also data acquisition, error analysis, deployment, and many others. @DillonLaird and I wrote this blog on how to design an ideal imaging system for machine vision
The first step in collecting training #data for #MachineVision is to design your image acquisition system. Here's our advice on creating quality images! https://xitter.gq/t.co/0rGzUAF7Fd #ComputerVision
Dillon Laird retweeted
Announcing our first international office in Medellin, Colombia! Landing AI, deeplearning.ai and AI Fund have teams there. AI needs to expand beyond a small handful of hubs like Silicon Valley and Beijing, and I’m bullish about Latin America. medium.com/@andrewng/ai-in-l…
Dillon Laird retweeted
EfficientNets: a family of more efficient & accurate image classification models. Found by architecture search and scaled up by one weird trick.
Link: arxiv.org/abs/1905.11946
Github: https://xitter.gq/t.co/vrRnVJVQNh
Blog: https://xitter.gq/t.co/0pPZmTxfUs
Dillon Laird retweeted
Introducing MobileNetV3: Based on MNASNet, found by architecture search, we applied additional methods to go even further (quantization friendly SqueezeExcite & Swish + NetAdapt + Compact layers). Result: 2x faster and more accurate than MobileNetV2. Link: arxiv.org/abs/1905.02244
Dillon Laird retweeted
New result from our @Stanford lab! Instead of giving everyone the same treatment, you can use ML to predict who would benefit from high blood pressure treatment and who wouldn't. @tonyduan_ @pranavrajpurkar @DillonLaird, and Sanjay Basu. Details here: ahajournals.org/doi/full/10.… – at Stanford, CA
Dillon Laird retweeted
JAX: GPU- and TPU-backed #NumPy with differentiation and JIT compilation. github.com/google/jax via @fonnesbeck

























