Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Learning to Reason with LLMs

openai.com

Post reply on HN
Print view View original

991–1000 of 1001 posts

Re: Learning to Reason with LLMs

#991

Post by OutOfHere »

delusional wrote: post #959
OutOfHere wrote: post #869

Earlier quoted context omitted.

> What's even are you? An accelerator of reaching the Singularity. This is something more.

You realize that you're not going inside the computer right? At best you're going to create a simulacrum of you. Something that looks, talks, and acts like you. It's never going to actually be you. You're going to be stuck out here with the rest of us, in whatever world we create in pursuit of the singularity suicide cult.

My friend, it has nothing to do with going inside a computer. Do not confuse the Singularity with mind uploading which is a distinct concept. The singularity has to do with technology acceleration, and with the inability to predict what lies beyond it. As such, it has nothing to do with any suicide cult. Please stop spreading nonsense about it. I do care about life in the physical world, not about a digital life.

Re: Learning to Reason with LLMs

#992

Post by anigbrowl »

infecto wrote: post #899
anigbrowl wrote: post #836

Earlier quoted context omitted.

you need to be on their tier 5 level, which requires $1,000 total paid and [...] Good opening for OpenAI's competitors to run a 'we're not snobs' promotion.

How so? I think most of the competition does this. Early partners/heavy users get access first which 1) hopefully provides feedback on the product and 2) provides a mechanism to stagger the release.

Marketing is about feelings, not facts.

Re: Learning to Reason with LLMs

#993

Post by diedyesterday »

packetlost wrote: post #191
ARandumGuy wrote: post #137

One thing that makes me skeptical is the lack of specific labels on the first two accuracy graphs. They just say it's a "log scale", without giving even a ballpark on the amount of time it took. Did the 80% accuracy test results take 10 seconds of compute? 10 minutes? 10 hours? 10 days? It's impossible to say with the data they've given us. The coding section indicates "ten hours to solve six challenging algorithmic…

When one axis is on log scale and the other is linear with the plot points appearing linear-ish, doesn't it mean there's a roughly exponential relationship between the two axis?

A linear graph with a log scale on the vertical axis means the original graph had near exponential growth.

A linear graph with a log scale on the the horizontal axis means the original graph had law of diminishing return kick it (somewhat similar to logarithmic but with a vertical asymptote).

Re: Learning to Reason with LLMs

#994

Post by deeznuttynutz »

jstummbillig wrote: post #600
cs391231 wrote: post #408

Student here. Can someone give me one reason why I should continue in software engineering that isn't denial and hopium?

One angle: There are a million SMBs and various other institutions, using none or really shitty software, that could be xx% to xxx% times more productive with custom software that they would never have been able to afford before. Now they can, en masse, because you will be able to built it a lot faster. I have been coding a lot with AI recently. Understanding and putting into thought what is needed for the program to…

This 1000%

Re: Learning to Reason with LLMs

#996

Post by diedyesterday »

utdiscant wrote: post #326

Feels like a lot of commenters here miss the difference between just doing chain-of-thought prompting, and what is happening here, which is learning a good chain of thought strategy using reinforcement learning. "Through reinforcement learning, o1 learns to hone its chain of thought and refine the strategies it uses." When looking at the chain of thought (COT) in the examples, you can see that the model employs diffe…

Reminds me of how Google's AlphaGo learned to play the best Go that was ever seen. And this somewhat seems a generalization of that.

Re: Learning to Reason with LLMs

#997

Post by trustno2 »

vanviegen wrote: post #529
trustno2 wrote: post #458

Earlier quoted context omitted.

That is in chatgpt now and it greatly improves chatgpt. What are you on to now?

Audio has only rolled out to a small subset of paying customers. There's still no word about the direct-from-4o image generation they demo'd. Let alone the video capabilities. So no, it's not in chatgpt.

ah okay you got a point

Re: Learning to Reason with LLMs

#998

Post by niutech »

tarruda wrote: post #605
seydor wrote: post #93

Earlier quoted context omitted.

The open source/weights models so far have proved that openAI doesn't have some special magic sauce. I m confident we ll soon have a model from Meta or others that s close to this level of reasoning. [Also consider that some of their top researchers have departed] On a cursory look, it looks like the chain of thought is a long series of chains of thought balanced on each step, with a small backtracking added whenever…

I would love to see Meta releasing CoT specialized model as a LoRa we can apply to existing 3.1 models

Isn't it what Reflection 70b (https://news.ycombinator.com/item?id=41459781) does on top of Llama 3.1?

Re: Learning to Reason with LLMs

#999

Post by schappim »

kenschu wrote: post #888
schappim wrote: post #885

If you’re using the API and are on tier 4, don’t bother adding more credits to move up to tier 5. I did this, and while my rate limits increased, the o1-preview / o1-mini model still wasn’t available.

Keep us posted!

I can confirm that the following models have since come in:

  • o1-preview-2024-09-12
  • o1-preview
  • o1-mini-2024-09-12
  • o1-mini

Re: Learning to Reason with LLMs

#1000

Post by AI_beffr »

mewpmewp2 wrote: post #923
AI_beffr wrote: post #797

Earlier quoted context omitted.

this is completely illogical. this is like gambling your life savings and as the die are rolling you say "i am incredibly fortunate to be witnessing this." like, you need to know the outcome before you know whether it was fortunate or unfortunate... this could be the most unfortunate thing that has ever happened in history.

I don't think we are anywhere near this, but if we were - whatever happens, happens, right. There is no escape from that.

you could apply this logic to global warming, pollution, child molestation. i mean youre so wrong its not even worth getting in to.
Post reply on HN