Soheil Bahrampour

Machine Learning/ Engineering Manager @ Google, prev: Deep Learning @ Bosch Research, PSU Machine Learning PhD

[Soheil Bahrampour @soheilbp]

@AlanEyre1 How low you democrats are willing to go, just to oppose Trump? Picking iran’s side over your president..wow

[Soheil Bahrampour @soheilbp]

#ThankYouTrump

[Soheil Bahrampour @soheilbp]

@realDonaldTrump Thank you Mr President for keeping your words!!

[Soheil Bahrampour @soheilbp]

@AOC Where is your statement on Iran regime brutality, slaughtering its people? Did you see the number of people killed just in 2 days? 32k+ was the number reorted by your government. How do you keep your mouth shut here while being so outspoken on Gaza? Why the double standard?

[Soheil Bahrampour @soheilbp]

@RoKhanna @RepThomasMassie Iranians in tech will unseat you. Where is your statement/resolution on condemning violanace happening in Iran?

[Soheil Bahrampour @soheilbp]

@BarackObama Any statement on Iran?

[Soheil Bahrampour @soheilbp]

@PavolRusnak @ZoubinGhahrama1

[Andrej Karpathy @karpathy]

In era of pretraining, what mattered was internet text. You'd primarily want a large, diverse, high quality collection of internet documents to learn from.

In era of supervised finetuning, it was conversations. Contract workers are hired to create answers for questions, a bit like what you'd see on Stack Overflow / Quora, or etc., but geared towards LLM use cases.

Neither of the two above are going away (imo), but in this era of reinforcement learning, it is now environments. Unlike the above, they give the LLM an opportunity to actually interact - take actions, see outcomes, etc. This means you can hope to do a lot better than statistical expert imitation. And they can be used both for model training and evaluation. But just like before, the core problem now is needing a large, diverse, high quality set of environments, as exercises for the LLM to practice against.

In some ways, I'm reminded of OpenAI's very first project (gym), which was exactly a framework hoping to build a large collection of environments in the same schema, but this was way before LLMs. So the environments were simple academic control tasks of the time, like cartpole, ATARI, etc. The @PrimeIntellect environments hub (and the `verifiers` repo on GitHub) builds the modernized version specifically targeting LLMs, and it's a great effort/idea. I pitched that someone build something like it earlier this year:
https://t.co/ANHhasxzD8
Environments have the property that once the skeleton of the framework is in place, in principle the community / industry can parallelize across many different domains, which is exciting.

Final thought - personally and long-term, I am bullish on environments and agentic interactions but I am bearish on reinforcement learning specifically. I think that reward functions are super sus, and I think humans don't use RL to learn (maybe they do for some motor tasks etc, but not intellectual problem solving tasks). Humans use different learning paradigms that are significantly more powerful and sample efficient and that haven't been properly invented and scaled yet, though early sketches and ideas exist (as just one example, the idea of "system prompt learning", moving the update to tokens/contexts not weights and optionally distilling to weights as a separate process a bit like sleep does).

[AK @_akhaliq]

ChatGPT or Grammarly? Evaluating ChatGPT on Grammatical Error Correction Benchmark

abs: https://t.co/63hSOZrQx6

[Soheil Bahrampour @soheilbp]

@bneyshabur @DeepMind @OriolVinyalsML Congrats!

[Scott Condron @_ScottCondron]

What actually happens when you call .backwards() in @PyTorch?

Autograd goodness 🪄!

PyTorch keeps track of all of the computations you’ve done on each of your tensors and .backwards() triggers it to compute the gradients and stores them in .grad.

1/3

[Urs Hölzle @uhoelzle]

If you need to see something uplifting for a change, and have 5 minutes, watch this. "It's literally the first time I've run free for...decades."

https://t.co/5jIktDKOtm

[Soheil Bahrampour @soheilbp]

@ElectProject Thank you! Amazing stats!

[Sundar Pichai @sundarpichai]

Since our last update, Google has accelerated our work to help people stay safe, informed & connected. Here’s the latest in our ongoing global #COVID19 response. We’ll keep doing everything we can to help people take care of themselves & their communities
https://t.co/a9fPAJIzuY

[Julien Mairal @julienmairal]

Video of my moment of glory at @icmlconf 2019
https://t.co/ffrmIYQmeA
and the slides
https://t.co/vTtf0iDiOL. A great thanks to @kamalikac and @rsalakhu (and many other people) for the time and efforts they devoted to the community when organizing such a large event.

[Soheil Bahrampour @soheilbp]

@bneyshabur @ylecun @prfsanjeevarora Congrats!

[TensorFlow @TensorFlow]

We’re excited to share that weight pruning is now part of the TensorFlow Model Optimization Toolkit!

Learn how sparsity can dramatically reduce model sizes with negligible accuracy loss.

Read here → https://t.co/bhOm3s6ixP

[Yann LeCun @ylecun]

Videos of all sessions at ICLR 2019 (including live stream)

https://t.co/sRFC7TLJuZ https://t.co/8VwjjQSanQ

[Soheil Bahrampour @soheilbp]

Attend the debate at #iclr2019 by going to https://t.co/TVS16h2jgX code: #iclrdebate

[Zico Kolter @zicokolter]

Our research group is starting a (technical) blog! First post, by @RICEric22, covers provable adversarial defenses https://t.co/TO9QYyAnOQ. Each post is downloadable as a Jupyter notebook, so you can recreate all the examples. More info about blog here https://t.co/qIBxeptqbB.