We show that a GPT-3 model can learn to express uncertainty about its own answers in natural language -- without use of model logits. When given a question, the model generates both an answer and a level of confidence (e.g. "90% confidence" or "high confidence"). These levels map to probabilities that are well calibrated. The model also remains moderately calibrated under distribution shift, and is sensitive to uncertainty in its own answers, rather than imitating human examples. To our knowledge, this is the first time a model has been shown to express calibrated uncertainty about its own answers in natural language. For testing calibration, we introduce the CalibratedMath suite of tasks. We compare the calibration of uncertainty expressed in words ("verbalized probability") to uncertainty extracted from model logits. Both kinds of uncertainty are capable of generalizing calibration under distribution shift. We also provide evidence that GPT-3's ability to generalize calibration depends on pre-trained latent representations that correlate with epistemic uncertainty over its answers.
Metadata: arXiv. Abstracts retain their original authors’ rights.
In the conversation
Original HN comments, with labels from simple text rules. A label describes the whole comment and may miss context or irony. Read the discussion to make up your own mind. About the labels
Related:
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets - https://arxiv.org/abs/2310.06824
The Internal State of an LLM Knows When It's Lying - https://arxiv.org/abs/2304.13734
LLMs Know More Than What They Say - https://arjunbansal.substack.com/p/llms-know-more-than-what-...
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
>I know my world model is fundamentally incomplete. Even more foundationally, I know that there is a world, and when my world model and the world disagree, the world wins.
Yeah this isn't really true. There's not how humans work. For a variety of reasons, Plenty stick with their incorrect model despite the world indicating otherwise. In fact, this seems to be normal enough human behaviour. Everyone does it, for something or the other. You are no exception.
And yes LLMs can in fact tell truth from fiction.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models…
Read the full comment
Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets - https://arxiv.org/abs/2310.06824
Your argument seems to boil down to "they can't perform experiments" but that isn't true either.
> If GPT-6 was only larger, it would still only know the things in its training set, whereas a human can take arbitrary time to think, can go out and perform experiments, can Google/GPT things themselves, etc.
A GPT model already has many ways of thinking and self-distilling; for example, it can give itself a lot more time to think using 'inner monologue' techniques, which make it capable of calculating out step by step, using external tools like Python etc: https://www.gwern.net/docs/ai/gpt/inner-monologue/index A GPT model can also already Google things, see WebGPT https://openai.com/blog/webgpt/ (itself merely one entry in a rapidly-expanding area of retrieval models…
Read the full comment
https://www.gwern.net/docs/ai/retrieval/index ). Enabling them to use web tools is likewise an active and increasingly successful area of research https://arxiv.org/abs/2202.08137#deepmind , with startups forming for that purpose: https://www.adept.ai/post/introducing-adept It can't 'run experiments', but the setting of reinforcement learning would allow that, and Decision Transformer (literally just a GPT trained on RL data and then given a prompt) is one of the more exciting directions - you may have seen Gato https://www.deepmind.com/publications/a-generalist-agent but have you also seen Multi-Game Transformer https://sites.google.com/view/multi-game-transformers ? Oh, that's just a single agent and you're not worried? Well that's fine because DT is so general it's almost trivial to make it do multi-agent RL as well https://arxiv.org/abs/2112.02845 https://arxiv.org/abs/2205.14953
Things are moving fast, and few people know all of the things that a GPT can do. (By the way, did you know that this 'big blob of inanimate knowledge' also knows what it doesn't know and can give well-calibrated predictions of accuracy https://arxiv.org/abs/2205.14334 ?)
Pretty much all research (and there's a fair few with different methodologies) on this converge on the same conclusion:
LLMs internally know a lot more about the uncertainty and factualness of their predictions than they say. "LLMs are always hallucinating" is a popular stance but wrong all the same. Maybe rather than asking Why models hallucinate, the better question is to ask "Why not?". During pre-training, there's close to zero incentive to push any uncertainty to the forefront (words).
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
The Geometry of Truth: Emergent Linear…
Read the full comment
Structure in Large Language Model Representations of True/False Datasets - https://arxiv.org/abs/2310.06824
The Internal State of an LLM Knows When It's Lying - https://arxiv.org/abs/2304.13734
LLMs Know More Than What They Say - https://arjunbansal.substack.com/p/llms-know-more-than-what-...
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
That's not true. There's been several papers probing this with different methodologies and the conclusion is pretty clear.
LLMs know a whole lot more about the uncertainty of their predictions than they say.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets - https://arxiv.org/abs/2310.06824
The Internal State of an LLM Knows When It's Lying - https://arxiv.org/abs/2304.13734
LLMs Know More Than What They Say - https://arjunbansal.substack.com/p/llms-know-more-than-what-...
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
>An LLM can't tell the difference between fact and fiction, because it can't apply logic.
Not true. They can differentiate it just fine. Of course being able to tell the difference and being incentivized to communicate it are 2 different things.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets - https://arxiv.org/abs/2310.06824
There really is no difference. Either there is utility or there isn't. "Fake" intelligence that produces results is just something that does not make any sense.
LLMs may be worse but Humans confidently state something they are uncertain about all the time lol. Maybe not experts in general but then that's still comparing LLMs to a small percentage of humanity.
At any rate, Unlike many seem to think, The issue is not in fact the lack of ability to distinguish truth and fact from fiction. Turns out being able to distinguish the two and having the incentive to communicate that are 2 different things.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for…
Read the full comment
Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets - https://arxiv.org/abs/2310.06824
>The human model isn’t just trained in language though, it’s trained on the world.
It's not trained on "the world". It's trained on a small slice of it, a few senses that are themselves slimmed down and then fabricated at parts.
To the bird that can feel and sense electromagnetic waves intuitively to guide travels, you're not trained on "the world"
>This is why they are so prone to hallucination.
No it's not. LLMs can distinguish between truth and hallucination. They just don't care to communicate that.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models…
Read the full comment
Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets - https://arxiv.org/abs/2310.06824
>More importantly, this is because it doesn't understand the meaning of what it generates, just juggles weights.
Yeah that's not the reason at all.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets - https://arxiv.org/abs/2310.06824
>But the results are indicative of what it is, and they fall short even if your only criteria is usefulness
There isn't a testable definition of general intelligence that GPT-4 fails that a chunk of humans also wouldn't.
If they fall short then so do people. Guess those people aren't generally intelligent then according to you.
>It can't tell the difference between truth and falsity
Yes, it very much can.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their…
Read the full comment
Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets - https://arxiv.org/abs/2310.06824
>The more interesting question is why are people like you personally offended
I'm not offended about anything.
Related
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
Language models (the computation) "care" about making "correct" predictions (correct of course being determined by the data).
This is why the simplest solution to reducing hallucinations is just making them more competent/better predictors (more neurons, more data). If the prediction can benefit from all the knowledge/reasoning it has accumulated from training, it'll tap that first. One of the biggest tells of hallucinations is if the model is saying something completely different each generation vs being consistent.
What they just don't care about is communicating being out of distribution or making whack predictions. This becomes a problem for humans because they're…
Read the full comment
perfectly fine making things up when the above fail.
But by all accounts, they do learn to distinguish these things. The computation is very much aware when it is going way off base.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets - https://arxiv.org/abs/2310.06824
Related -
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
>You are unable to.
This is just wrong lol.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
> Numbers (LLMs still can't do math and the ones that "can" are very brute forced and don't generalize)
These seem to generalize.
https://arxiv.org/abs/2211.09066
https://arxiv.org/abs/2308.00304
> Commutative properties (The (idk why anyone was surprised) "reversal curse": https://arxiv.org/abs/2309.12288)
This was a retrieval from training issue not an inference one. and make no mistake, people will fail these sort of answers too (perhaps not to the same extent). Very often someone will learn a fact in a certain why and fail to recall that when asked in a reverse or roundabout way.
Communicative properties are not something that occur in the vast majority of text or…
Read the full comment
language. It's not "generalization" for training to treat it as such lol. Since they can answer such questions just fine during inference, It makes little sense to say an LLM doesn't understand communicative properties.
> How to say "I don't know"
There's quite a lot of indication that the computation can distinguishing hallucinations. It just has no incentive to communicate this. Don't blame the LLM here. Blame the very human data that doesn't encourage this.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
I'm not trying to say that LLMs are perfect. and as long as people understand that saying an LLM doesn't understand x doesn't mean an LLM doesn't understand anything at all, i'll happily state weaknesses.
There's quite a lot of indication that the computation can distinguishing hallucinations. It just has no incentive to communicate this.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
>As a rule, chatbots today have a propensity to confidently make stuff up, or, as some researchers say, “hallucinate.” At the root of these hallucinations is an inability to introspect: the A.I. doesn’t know what it does and doesn’t know.
The last bit doesn't seem to be true. There's quite a lot of indication that the computation can distinguishing hallucinations. It just has no incentive to communicate this.
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their…
Read the full comment
Uncertainty in Words - https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know - https://arxiv.org/abs/2207.05221
Also even if we're strictly talking about text, there is still a ton of data left to train on. We've just barely reached what is easily scrapable online and are nowhere near a real limit yet. And of course, you can just train more than one epoch. That said, it's very clear quality data is far more helpful than sheer quantity and sheer quantity is more likely than not to derail progress.
Related -
GPT-4 logits calibration pre RLHF - https://imgur.com/a/3gYel9r
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback - https://arxiv.org/abs/2305.14975
Teaching Models to Express Their Uncertainty in Words -
https://arxiv.org/abs/2205.14334
Language Models (Mostly) Know What They Know -
https://arxiv.org/abs/2207.05221
Artificial intelligence (AI) researchers have been developing and refining large language models (LLMs) that exhibit remarkable capabilities across a variety of domains and tasks, challenging our understanding of learning and cognition.
Recent work has demonstrated substantial gains on many NLP tasks and benchmarks by pre-training on a large corpus of text followed by fine-tuning on a specific task.
The Fermi paradox is the conflict between an expectation of a high {\em ex ante} probability of intelligent life elsewhere in the universe and the apparently lifeless universe we in fact observe.
Despite their massive size, successful deep artificial neural networks can exhibit a remarkably small difference between training and test performance.