Discussion

Autoresearch on an old research idea

love2read: So... It did work. It found bugs (that he didn't know about) and it did optimization (that he hadn't done).

datsci_est_2015: I often use LLMs to explore prior art and maybe find some alternative ways of thinking of problems. About 90% of what it tells me is useless or inapplicable to my domain due to a technicality it could not have known, but the other 10% is nice and has helped me learn some great new things.I can’t imagine letting an agent try everything that the LLM chatbot had recommended ($$$). Often coming up in recommendations are very poorly maintained / niche libraries that have quite a lot of content written about them but what I can only imagine is very limited use in real production environments.On the other hand, we have domain expert “consultants” in our leadership’s ears making equally absurd recommendations that we constantly have to disprove. Maybe an agent can occupy those consultants and let us do our work in peace.

MattGaiser: > agent try everything that the LLM chatbot had recommended ($$$)A lot depends on whether it is expensive to you. I use Claude Code for the smallest of whims and rarely run out of tokens on my Max plan.

BrokenCogs: Does autoresearch work for projects that are not llm based? Eg in karpathy's example he is optimizing the nanogpt. What if I wanted to improve a Unet for image segmentation?

Eufrat: I find LLMs useful in regurgitating one-liners that I can’t be bothered to remember or things where even being flat out wrong is okay and you just do it yourself.For all the folks spending a lot of time and energy in setting up MCP servers, AGENTS.md, etc. I think this represents more that the LLM cannot do what it is being sold as by AI boosters and needs extreme amounts of guidance to reach a desired goal, if it even can. This is not an argument that the tech has no value. It clearly can be useful in certain situations, but this is not what OpenAI/Anthropic/Perplexity are selling and I don’t think the actual use cases have a sustainable business model.People who spend the energy to tailor the LLMs to their specific workflows and get it to be successful, amazing. Does this scale? What’s going to happen if you don’t have massive amounts of money subsidizing the training and infrastructure? What’s the actual value proposition without all this money propping it up?

foobarian: > I find LLMs useful in regurgitating one-liners that I can’t be bothered to rememberI found LLMs make a fabulous frontend for git :-D

bethekind: I used it to speed up an codecompass-like repo from 86 files per second to 2000. Still haven't used the repo in production, so maybe it secretly broke things, but the ability to say: "optimize this benchmark and commit only if you pass these tests" is nice

jpcompartir: There are better techniques for hyper-parameter optimisation, right? I fear I have missed something important, why has Autoresearch blown up so much?The bottleneck in AI/ML/DL is always data (volume & quality) or compute.Does/can Autoresearch help improve large-scale datasets? Is it more compute efficien than humans?

hun3: [delayed]

simonw: Tobi from Shopify used a variant of autoresearch to optimize the Liquid template engine, and found a 53% speedup after ~120 experiments: https://github.com/Shopify/liquid/pull/2056I wrote up some more notes on that here: https://simonwillison.net/2026/Mar/13/liquid/

Denzel: How much did this cost? Has there ever been an engineering focus on performance for liquid?It’s certainly cool, but the optimizations are so basic that I’d expect a performance engineer to find these within a day or two with some flame graphs and profiling.

_pdp_: Take some working code. Ask an LLM to fix bugs. Measure performance and test coverage. Feed the results back into the LLM. Repeat.This has been the standard approach for more complex LLM deployments for a while now in our shop.Using different models across iterations is also something I've found useful in my own experiments. It's like getting a fresh pair of eyes.

cyanydeez: Can we modify this approach to get LLMs that are good at specific programming languages or frameworks? That seems to be where local LLMs could really shine.

simonw: He used Pi as the harness but didn't say which underlying model. My stab-in-the-air guess would be no more than a few hundred dollars in token spend (for 120 experiments run over a few days assuming Claude Opus 4.6 used without the benefits of the Claude Max plan.)So cheaper than a performance engineer for a day or two... but the Shopify CEO's own time is likely a whole lot more expensive than a regular engineer!

nextos: AFAIK, it's a bit more than hyper-parameter tuning as it can also make non-parametric (structural) changes.Non-parametric optimization is not a new idea. I guess the hype is partly because people hope it will be less brute force now.

gwerbin: It's an LLM-powered evolutionary algorithm.

lucasay: This feels less like automated research and more like structured trial and error with a decent feedback loop. Still useful, but I think the real bottleneck is how good your eval metric is. If that’s weak, the whole loop just optimizes for the wrong thing faster.

kridsdale1: I mean, isn’t that “the scientific method”?

lamroger: Awesome breakdown! It really feels like a hyper-hyper parameter search + bug fixer.I started looking at Kaggle again and autoresearch seems to converge to many of the solution vibes there.Wild ensembles, squeezing a bit of loss out. More engineering than research IMO

sdenton4: For raw hyperparameter search, though, I would expect a proper Bayesian framework to be much better. Eg, vizier.

ainch: I think it depends whether you can leverage some knowledge. It's possible for a person/LLM to look at a loss curve and say "oh that's undertraining, let's bump the lr" - whereas a Bayesian method doesn't necessarily have deeper understanding, so it'll waste a lot of time exploring the search space on poor options.If you're resource unconstrained then BO should ofc do very well though.

ainch: I'd like see a system like this take more inspiration from the ES literature, similar to AlphaEvolve. Let's see an archive of solutions, novelty scoring and some crossover rather than purely mutating the same file in a linear fashion.

motbus3: I've done something with a small project I have and I had very similar results overall.

1970-01-01: > The original paper used several medical X-ray datasets which I don’t have access to anymore, so I needed a new dataset with spatial annotations to test the expert attention mechanism. I picked the Ukiyo-eVG dataset: ~11K Japanese woodblock printsThat's such a weird switch. There's lots of free medical imaging online. Example: https://www.cancerimagingarchive.net/

ks2048: I think image segmentation is in the same class as LLMs - ML experiments.What about more distant software projects? Give it the CPython source code and say you want it to be faster.