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

Learning from context is harder than we thought

hy.tencent.com

Post reply on HN
Print view View original

51–60 of 140 posts

Re: Learning from context is harder than we thought

#51

Post by hyperpape »

bradfa wrote: post #2

The key seems to be that you take the transcript of a model working within a problem domain that it’s not yet good at or where the context doesn’t match it’s original training and then you continually retrain it based on its efforts and guidance from a human or other expert. You end up with a specialty model in a given domain that keeps getting better at that domain, just like a human. The hard part is likely when so…

> At scale humans accept new things slowly.

I think this is true, but there are big differences. Motivated humans with a reasonable background learn lots of things quickly, even though we also swim in an ocean of half-truths or outdated facts.

We also are resistant to certain controversial ideas.

But neither of those things are really that analogous to the limitations on what models can currently learn without a new training run.

Re: Learning from context is harder than we thought

#52

Post by rabbitlord »

cs702 wrote: post #24

The problem is even more fundamental: Today's models stop learning once they're deployed to production. There's pretraining, training, and finetuning, during which model parameters are updated. Then there's inference, during which the model is frozen. "In-context learning" doesn't update the model. We need models that keep on learning (updating their parameters) forever, online, all the time.

I think they can do in-context learning.

Re: Learning from context is harder than we thought

#53

Post by reliabilityguy »

numtel wrote: post #16
Tor3 wrote: post #14

Earlier quoted context omitted.

"There are examples of tribal humans not being able to perceive a green square among blue squares, because their language does not have a word for the green color. Similarly, some use the same word for blue and white, and are unable to perceive them as different colors." Both of the above is false. There are a ton of different colors that I happen to call "red", that does not mean that I can't perceive them as differ…

There's a Terence McKenna quote about this: > So, for instance, you know, I’ve made this example before: a child lying in a crib and a hummingbird comes into the room and the child is ecstatic because this shimmering iridescence of movement and sound and attention, it’s just wonderful. I mean, it is an instantaneous miracle when placed against the background of the dull wallpaper of the nursery and so forth. But, the…

Very poetic, I like it.

Re: Learning from context is harder than we thought

#54

Post by fph »

4b11b4 wrote: post #31
cs702 wrote: post #24

The problem is even more fundamental: Today's models stop learning once they're deployed to production. There's pretraining, training, and finetuning, during which model parameters are updated. Then there's inference, during which the model is frozen. "In-context learning" doesn't update the model. We need models that keep on learning (updating their parameters) forever, online, all the time.

I'm not sure if you want models perpetually updating weights. You might run into undesirable scenarios.

Tay the chatbot says hi from 2017.

Re: Learning from context is harder than we thought

#55

Post by snovv_crash »

johnsmith1840 wrote: post #3

It's basically continual learning. This is beyond a hard problem it's currently an impossible one. I know of no system that solve CL even at small scale let alone large models. Annoyingly, they have SOME inherent capability to do it. It's really easy to get sucked down this path due to that glimmer of hope but the longer you play with it the more annoying it becomes. SSI seems to be focused on this problem directly s…

For neural networks, yeah continuous learning is basically dead.

But for other ML approaches, it works really well. KNN is one example that works particularly well.

Re: Learning from context is harder than we thought

#56

Post by jkhdigital »

joriJordan wrote: post #6

Because we don't experience reality through language but direct sensory perception. Language is arbitrary bird song and visual representations dragged forward from history, accepted definitions never uniformly distributed. Testing based on contextual correctness makes no sense when there is no center to the universe. No "one true context to rule them all". We learn from hands on sensory experiences. Our bodies store…

Your last statement misses the mark—of course the brain is the root of human intelligence. The error is in assuming that consciousness is the primary learning modality. Or, as you put it, “arguing semantics”.

From my own personal experience, this realization came after finally learning a difficult foreign language after years and years of “wanting” to learn it but making little progress. The shift came when I approached it like learning martial arts rather than mathematics. Nobody would be foolish enough to suggest that you could “think” your way to a black belt, but we mistakenly assume that skills which involve only the organs in our head (eyes, ears, mouth) can be reduced to a thought process.

Re: Learning from context is harder than we thought

#57

Post by kikoreis »

> Without any context provided, the state-of-the-art model, GPT-5.1 (High), is only able to solve less than 1% of tasks. This starkly demonstrates that the data is contamination-free, as the model is almost entirely incapable of solving the tasks without learning from the context.

[...]

[With context provided,] on average, models solve only 17.2% of tasks. Even the best-performing model, GPT-5.1 (High), achieves just 23.7%.

Re: Learning from context is harder than we thought

#58

Post by Legend2440 »

snovv_crash wrote: post #55
johnsmith1840 wrote: post #3

It's basically continual learning. This is beyond a hard problem it's currently an impossible one. I know of no system that solve CL even at small scale let alone large models. Annoyingly, they have SOME inherent capability to do it. It's really easy to get sucked down this path due to that glimmer of hope but the longer you play with it the more annoying it becomes. SSI seems to be focused on this problem directly s…

For neural networks, yeah continuous learning is basically dead. But for other ML approaches, it works really well. KNN is one example that works particularly well.

Ehhh KNN doesn’t have a training phase, so it’s really more that the concept of continual learning doesn’t apply. You have to store your entire dataset and recalculate everything from scratch every time anyway.

Re: Learning from context is harder than we thought

#59

Post by threecheese »

raincole wrote: post #50
cs702 wrote: post #24

The problem is even more fundamental: Today's models stop learning once they're deployed to production. There's pretraining, training, and finetuning, during which model parameters are updated. Then there's inference, during which the model is frozen. "In-context learning" doesn't update the model. We need models that keep on learning (updating their parameters) forever, online, all the time.

> We need models that keep on learning (updating their parameters) forever, online, all the time. Yeah, that's the guaranteed way to get MechaHilter in your latent space. If the feedback loop is fast enough I think it would finally kill the internet (in the 'dead internet theory' sense). Perhaps it's better for everyone though.

Many are working on this, as well as in-latent-space communication across models. Because we can’t understand that, by the time we notice MechaHitler it’ll be too late.

Re: Learning from context is harder than we thought

#60

Post by jkhdigital »

winddude wrote: post #26
godelski wrote: post #19

It is weird to read because they bring up many things a lot of people have been critiquing for years. > But as impressive as these feats are, they obscure a simple truth: being a "test-taker" is not what most people need from an AI. > In all these cases, humans aren't relying solely on a fixed body of knowledge learned years ago. We are learning, in real-time, from the context right in front of us. > To bridge this g…

> But as impressive as these feats are, they obscure a simple truth: being a "test-taker" is not what most people need from an AI. People have been bringing that up long before AI, on how schooling often tests on memorization and regurgitation of facts. Looking up facts is also a large part of the internet, so it is something that's in demand, and i believe a large portion of openAI/cluade prompts have a big overlap…

For all the disparagement of “fact regurgitation” as pedagogical practice, it’s not like there’s some proven better alternative. Higher-order reasoning doesn’t happen without a thorough catalogue of domain knowledge readily accessible in your context window.
Post reply on HN