YHK - yet another hacker news clone

Xiami Mimo 2.6 Live Post-Training Dashboard

60 points krackers 57 minutes ago 21 comments

wolttam 24 minutes ago | parent

Hah, it would be great to see more labs pick this up.

krm01 22 minutes ago | parent

This is pretty neat. What would be a good reason for the other Model providers to not do this?

kibae 12 minutes ago | parent

Speculating here, but I assume researchers can make a reasonable estimate of the size of closed models based on factors like training time, training speed, and the number of tokens processed.

Also, Anthropic and OpenAI probably want to keep each other on their toes so they don’t end up on the wrong side of another Opus 4.6 / GPT-5.3-Codex situation, where one lab releases a model only for the other to drop a better one hours later.

speedgoose 19 minutes ago | parent

I didn't know 2 thirds of the training data would be source code.

jerrygenser 13 minutes ago | parent

that is the the "data used to improve the model" when signing up for the subscription plans

leothetechguy 13 minutes ago | parent

this is the rl run, not the pretraining run

thehamkercat 15 minutes ago | parent

This is crazy, but sadly anthropic/openai will never do this, what has happened to this world, where chinese companies are more open than US or even EU companies

rozab 14 minutes ago | parent

Why are they doing this? To try head off accusations about distillation?

bayindirh 7 minutes ago | parent

Sometimes you're confident about what you're doing and show how you work to the world.

Keeping the garage door open, or at least making the door translucent. It's always cool.

liuliu 14 minutes ago | parent

When you run benchmarks while training, isn't that the definition of contamination? Asking because I am not sure if this is normal in big labs now.

lucrbvi 11 minutes ago | parent

They are using it to evaluate checkpoints during the training, they are probably not using the benchmarks for training the models. It's a common practice for big reinforcement learning runs.

jampekka 6 minutes ago | parent

Kinda yes. The benchmarks become part of the validation set, which means the models get slightly overfit to them if they are used as criteria for stopping the training. But a lot less compared to using them in the training data.

I'd guess everybody uses at least some benchmarks as stopping criteria, which is kinda sensible, but it also does induce some benchmaxxing, and explains partly why the newest models always tend to eke out in benchmarks.

https://en.wikipedia.org/wiki/Training,_validation,_and_test...

liuliu 2 minutes ago | parent

Correct. If just stopping criteria, that is less contaminated. The question gets muddier once you also use it to determine hyperparameters during small-scale runs.

SwellJoe 4 minutes ago | parent

You gotta have something to aim at. And, presumably, the benchmark is not part of the training data, it is the test against which the model is tested at each stage; is behavior moving in the right direction?

ProfessorLayton 13 minutes ago | parent

2.6 Pro: >started 2026-09-15 10:32 UTC

For some reason I thought training took much, much longer than what the progress bar suggests.

This is really neat, I'm currently using mimo 2.5 pro, and it's decent (or great given the price). Hopefully their next one is multimodal.

GaggiX 8 minutes ago | parent

These are post-training reinforcement learning steps.

krackers 6 minutes ago | parent

Yes, updated the submission title to say "post-training" to hopefully prevent further confusion

joelwallis 12 minutes ago | parent

I been using MiMo-V2.5 to do most of my work as software engineer, on a variety of projects I'm working on, and I been VERY happy with ROI. The model is very powerful! Not perfect – I've run in hallucination loops once or twice, but nothing a stop-then-continue wouldn't solve.

The cost is unbelievably low, and the quality of intelligence I get is equivalent to when I was working mostly with Anthropic models (late last year/early this year). I'm fully invested in MiMo and I'm very happy with it.

-- PS: I also check almost daily to see if other models are capable of doing such great work. And they do – DS4F is powerful and DS41 is impressive, GLM 5.3 Flash gets a job done well, etc. – but when I add cost of M-token in the ROI math, Jeez! MiMo is an order of magnitude better.

james2doyle 8 minutes ago | parent

2.5 Pro or the regular 2.5?

I always found that those Mimo models to be really good at tool calling and following instructions

levocardia 9 minutes ago | parent

You'd think they would make it less obvious that they are running their whole operation with Claude

SwellJoe 6 minutes ago | parent

It's not obvious to me. What's the tell?