Conversation

Notices

Embed this notice
Prof. Emily M. Bender(she/her) (emilymbender@dair-community.social)'s status on Thursday, 25-May-2023 17:01:11 JST Prof. Emily M. Bender(she/her)

A thought experiment in the National Library of Thailand—or why #ChatGPT (or any other language model) isn't actually understanding.
https://medium.com/@emilymenonbender/thought-experiment-in-the-national-library-of-thailand-f2bf761a8a83

In conversation Thursday, 25-May-2023 17:01:11 JST from dair-community.social permalink
- clacke likes this.
- Embed this notice
  pettter (pettter@mastodon.acc.umu.se)'s status on Thursday, 25-May-2023 17:01:10 JST pettter
  in reply to
  
  @emilymbender The obvious counter to this is that LLM are not strictly given _text_ but also _communications_ and in particular through RLHF _actual communications and feedback on its "utterances"_.
  I agree with the basic point that LLM are far more appearance than substance, and that people are very good at finding meaning where there is none, but there are pathways open for actual LLM as they exist or might exist that are not present in your example.
  
  In conversation Thursday, 25-May-2023 17:01:10 JST permalink
- Embed this notice
  clacke (clacke@libranet.de)'s status on Sunday, 28-May-2023 15:59:45 JST clacke
  in reply to
  - Lukas Galke
  - Andrew Lampinen
  @lpag Something unimaginable might be able to learn to understand the meaning behind language from just studying texts.
  But we are the ones who imagined the LLM.
  @lampinen @emilymbender
  
  In conversation Sunday, 28-May-2023 15:59:45 JST permalink
- Embed this notice
  Lukas Galke (lpag@sigmoid.social)'s status on Sunday, 28-May-2023 15:59:51 JST Lukas Galke
  in reply to
  - Andrew Lampinen
  @emilymbender I wonder if, in general, it is fair to conclude that only because it is not imaginable that we can do something, some other learning system (as an LLM) cannot do the thing?
  On related note: I would love to hear your take on @lampinen et al.'s recent work: https://arxiv.org/abs/2305.16183
  In conversation Sunday, 28-May-2023 15:59:51 JST permalink
  Attachments
  1. Domain not in remote thumbnail source whitelist: static.arxiv.org
    
    Passive learning of active causal strategies in agents and language models
    
    What can be learned about causality and experimentation from passive data? This question is salient given recent successes of passively-trained language models in interactive domains such as tool use. Passive learning is inherently limited. However, we show that purely passive learning can in fact allow an agent to learn generalizable strategies for determining and using causal structures, as long as the agent can intervene at test time. We formally illustrate that learning a strategy of first experimenting, then seeking goals, can allow generalization from passive learning in principle. We then show empirically that agents trained via imitation on expert data can indeed generalize at test time to infer and use causal links which are never present in the training data; these agents can also generalize experimentation strategies to novel variable sets never observed in training. We then show that strategies for causal intervention and exploitation can be generalized from passive data even in a more complex environment with high-dimensional observations, with the support of natural language explanations. Explanations can even allow passive learners to generalize out-of-distribution from perfectly-confounded training data. Finally, we show that language models, trained only on passive next-word prediction, can generalize causal intervention strategies from a few-shot prompt containing examples of experimentation, together with explanations and reasoning. These results highlight the surprising power of passive learning of active causal strategies, and may help to understand the behaviors and capabilities of language models.
- Embed this notice
  JW prince of CPH (jwcph@norrebro.space)'s status on Sunday, 28-May-2023 16:03:25 JST JW prince of CPH
  in reply to
  
  @emilymbender A Thai library, or a Chinese room... https://en.m.wikipedia.org/wiki/Chinese_room
  Also, reminds me of the time an astronomer thought he could see canals on Mars which appeared to be created by an intelligence - there aren't, of course, so another guy later observed that sure, the lines he thought he saw in his telescope were indeed signs of intelligence, but on the opposite end of the scope...
  In conversation Sunday, 28-May-2023 16:03:25 JST permalink
  Attachments
  1. Untitled attachment
  clacke likes this.

Public

Conversation

Notices

Feeds