[HN Gopher] Q-Transformer
       ___________________________________________________________________
        
       Q-Transformer
        
       Author : jonbaer
       Score  : 205 points
       Date   : 2023-11-30 10:42 UTC (12 hours ago)
        
 (HTM) web link (qtransformer.github.io)
 (TXT) w3m dump (qtransformer.github.io)
        
       | amelius wrote:
       | 33% success rate for opening a drawer?
        
         | varjag wrote:
         | That's a great start. Opening a drawer is pretty hard problem.
        
           | methodical wrote:
           | Correct- as a reference for OC this is known as Moravec's
           | Paradox (https://en.wikipedia.org/wiki/Moravec%27s_paradox)
        
             | varjag wrote:
             | Indeed: we have machines doing novel work in organic
             | synthesis but none yet to empty a dishwasher.
        
               | amelius wrote:
               | This already exists:
               | 
               | https://www.youtube.com/watch?v=UAG_FBZJVJ8
        
               | varjag wrote:
               | Could you point me to the timestamp where it empties a
               | dishwasher?
        
               | wiz21c wrote:
               | interacting with environment is the next frontier. See AI
               | driving...
        
             | pixl97 wrote:
             | At least to me the key point of the article
             | 
             | >Encoded in the large, highly evolved sensory and motor
             | portions of the human brain is a billion years of
             | experience about the nature of the world and how to survive
             | in it. The deliberate process we call reasoning is, I
             | believe, the thinnest veneer of human thought, effective
             | only because it is supported by this much older and much
             | more powerful, though usually unconscious, sensorimotor
             | knowledge. We are all prodigious olympians in perceptual
             | and motor areas, so good that we make the difficult look
             | easy. Abstract thought, though, is a new trick, perhaps
             | less than 100 thousand years old. We have not yet mastered
             | it. It is not all that intrinsically difficult; it just
             | seems so when we do it.
        
               | amelius wrote:
               | I'm not convinced. There's a lot of creative thinking
               | going on in evolutionary anthropology, but the proofs are
               | meager.
        
           | polygamous_bat wrote:
           | The other robot project posted here yesterday opened drawers
           | at around 80% though
           | https://news.ycombinator.com/item?id=38453047 And they did it
           | in many different homes.
        
             | sfink wrote:
             | My robot can open drawers at near 100%, but it takes the
             | form of a stick of dynamite.
             | 
             | Still working on closing the drawers afterwards, though...
        
               | coolspot wrote:
               | Perhaps you could automate rebuilding the drawers with
               | some IKEA/TaskRabbit APIs.
        
             | amelius wrote:
             | They should probably optimize for minimal friction force.
             | If you're pulling the handle not entirely in the direction
             | of movement, you're not doing it right.
        
       | luke-stanley wrote:
       | This came out a while ago, but I can see why it's been reposted.
       | It's another approach to searching the solution space with
       | rewards and transformer language models. Like XoT and others as I
       | pointed out in my last comment:
       | https://news.ycombinator.com/item?id=38471388
        
         | dartos wrote:
         | It's probably being reposted bc people think it has to do with
         | Q*
        
       | Philpax wrote:
       | Worth noting that this came from DeepMind in Sep-Oct of this
       | year, so it's not related to OpenAI's "unfortunate" leak [0] of
       | Q*.
       | 
       | [0] https://www.theverge.com/2023/11/29/23982046/sam-altman-
       | inte...
        
         | jansan wrote:
         | Also, it is not Amazon Q, which was presented yesterday:
         | 
         | https://aws.amazon.com/de/q/
        
           | 3cats-in-a-coat wrote:
           | Also it's not Q from Star Trek, or Q stars which are an
           | exotic form of matter immediately preceding the singularity
           | of a black hole.
        
             | GTP wrote:
             | Also, it's not Android Q.
        
               | kristjansson wrote:
               | Or James Bond's Q.
        
               | behnamoh wrote:
               | or Q.
        
         | px43 wrote:
         | I wouldn't go that far. Immediately after Q* leaked, Jimmy
         | Apples pointed to this thread on Twitter :
         | https://twitter.com/polynoamial/status/1676971508911198209
         | 
         | It's Noam Brown's (former Deep Mind) "I'm joining OpenAI"
         | thread from Back in June, and he talks about how he wants to
         | bring some of the work from AlphaGoZero etc into OpenAI's AGI
         | architecture.
         | 
         | It looks like this Q-transformer work shares a lot of the same
         | lineage of what Q* is supposed to be. This is Google's attempt
         | at mushing Q-learning into transformer land, which is probably
         | what Q* is as well. It's not the same thing, for sure, but they
         | are at least sibling bodies of work.
        
       | spacebacon wrote:
       | Operation reclaim the letter Q within a reasonable objective
       | reality.
        
         | VikingCoder wrote:
         | What's the letter?
         | 
         | What's the letter?
         | 
         | What's the letter?
         | 
         | The letter of the day is...
         | 
         | Q!
        
           | wiz21c wrote:
           | James Bond secret weapon ?
        
       | maurits wrote:
       | Being implemented as we speak, by the always impressive
       | LucidRains [1]
       | 
       | [1]: https://github.com/lucidrains/q-transformer
        
         | GaggiX wrote:
         | It's kinda funny how Lucidrains came back to make numerous
         | commits to this repo following Q*.
        
           | itissid wrote:
           | Apparently this guy like a bunch of others like
           | https://github.com/ggerganov/ggml are implementing
           | transformers from papers for people that want them. Pretty
           | cool.
        
             | milanove wrote:
             | I'm glad to see people in the open source community taking
             | the initiative to implement these tools for the general
             | public. It's one thing to understand the high-level
             | concepts in a paper, but another entirely to implement them
             | efficiently. Some people just really have a knack for
             | reading academic papers and implementing those ideas in
             | software.
        
             | pama wrote:
             | The additional link is indeed very useful, however, with
             | one exception, it is limited to inference, ie code for
             | evaluating a model using somebody else's trained model
             | weights. The original link has code for training (which can
             | also be used for inference, of course.) Inference is super
             | important, of course, so it makes a lot of sense to put
             | effort to make it work well on all kinds of hardware; the
             | papers typically focus on training and only then
             | evaluating/validating a new model.
        
         | rtev wrote:
         | Lucidrains also created the much-missed EpicMafia, which still
         | doesn't have a good replacement after shutting down.
         | Exceptionally skilled person!
        
       | brcmthrowaway wrote:
       | It's clear we're searching for the god algorithm of AI, just like
       | physicists are searching for theory of everything. Are
       | transformers the answer though?
        
         | mirekrusin wrote:
         | If they were we'd stop searching already, no?
        
         | goldenkey wrote:
         | There's really no God algorithm needed, just something good
         | enough to assist with research of the next tier of hardware,
         | energy, and code for AI.
        
         | Workaccount2 wrote:
         | It seems more likely that there will be multiple avenues to
         | AGI, all with their strengths and weaknesses. But perhaps the
         | "God AI" will be a multifaceted model composed of many
         | different models acting in unison.
        
           | falcor84 wrote:
           | Exactly, just as the true god has seven aspects [0].
           | 
           | [0] https://en.wikipedia.org/wiki/Themes_in_A_Song_of_Ice_and
           | _Fi...
        
             | bee_rider wrote:
             | On the other hand, most of the original gods were parts of
             | polytheistic pantheons. Maybe a bunch of models that
             | represent identities and biases could be more useful, they
             | could argue amongst themselves, presenting a more full
             | point of view, users could become familiar with the
             | particular perspectives.
        
           | kstrauser wrote:
           | I could see something like the "modularity of mind" model of
           | human consciousness, where multiple approaches are working on
           | "subconscious" solutions to a given problem in parallel, with
           | a top layer deciding which is appropriate at the moment.
           | 
           | My human brain doesn't use the same algorithm for learning to
           | play a song on a piano as learning to play a new board game.
           | I'm not an AI person, but it seems reasonable to imagine we'd
           | have different "modules" to apply as needed.
           | 
           | AlphaGo probably sucks at conversation. ChatGPT can't play
           | Go. The part of my brain writing this couldn't throw a
           | baseball. The physics engine that lets me throw a baseball
           | couldn't write this. Is there a reason we'd want or need one
           | specific AI approach to be universally applicable?
        
             | platz wrote:
             | The Bicameral Mind
        
           | two_in_one wrote:
           | Most likely there will be more then one specimen claiming to
           | be AGI. From different groups. And they will be hard to
           | compare.
        
         | theptip wrote:
         | No, not at all. Just looking for the next step of many,
         | stacking S-curves atop each other.
        
         | mikewarot wrote:
         | We're searching for an _efficient algorithm_ that leads to AGI.
         | Given sufficient time and compute, I 'm sure that we could get
         | there with existing stuff, by accident, and we wouldn't realize
         | it before moving on to the next thing... and there'd be a poor
         | orphan AGI, lost in a Git repo, waiting for runtime.
        
           | sfink wrote:
           | "Given sufficient time and compute" covers up a lot, though.
           | The ultimate God AGI that would be created through that sort
           | of process would take the form of a large room filled with a
           | whole lot of monkeys and typewriters.
        
         | lossolo wrote:
         | As it stands, each token processed by transformer requires a
         | constant amount of computation and energy. For an AGI system,
         | this would imply the ability to solve problems of any
         | complexity with a fixed amount of energy. But if this were
         | true, it would essentially mean that P equals NP, a major
         | theoretical breakthrough in computational complexity theory.
         | IMO we are still missing something.
        
         | AndrewKemendo wrote:
         | Algorithm is too simple, but yes - what does the system look
         | like
         | 
         | I think the transformers architecture, or something very
         | similar with eventually-on-policy time series forecasting in a
         | markov decision process, is the right answer actually and was
         | what I have been trying to make progress on for a long time[1].
         | 
         | [1]https://kemendo.com/research/streaminference.html
        
           | snovv_crash wrote:
           | That plus some Monte Carlo search as in AlphaZero makes a
           | very strong (although computationally heavy) contender for
           | 'alive' AI.
        
             | AndrewKemendo wrote:
             | After this past few years of progress on multivariate
             | stream forecasting I really think it's going to be
             | different implementations of the same basic streaming data
             | forecasting architecture
        
         | amelius wrote:
         | Theory of everything is supposed to be beautiful.
         | 
         | Theory of AI is just going to be some weird network with a
         | shit-ton of compute power, where the latter is more important
         | to the outcome than the former.
        
       | itissid wrote:
       | Could one of the roomba companies now make those things recognize
       | the objects and avoid them, instead of dead reckoning or RF based
       | obstacle avoidance?
        
       | itissid wrote:
       | Just to be clear the "scalable" part here is assuming the dataset
       | of human demonstrations is _available_ it can in a closed loop
       | learn better than before? In other words, am I right in
       | understanding that the mapping from Abstract actions to concrete
       | steps in the State space is provided at a low  "enough" cost?
        
         | itissid wrote:
         | Like in Workflow Guided explorations like:
         | https://arxiv.org/pdf/1802.08802.pdf, for example, if you said
         | "Forward the email" the Bot needs to sample from a set of
         | actions which in this case would be DOM elements and the JS API
         | to call to click the right button to forward the email. This
         | would be a harder/interesting to test out. Can it learn to
         | forward an email on firefox desktop browser and then do it on
         | an iphone?
         | 
         | It feels like opening a drawer once learn't, the bot has a
         | (world?) model of what all drawers might look like, so it can
         | open different drawers in any world. But this might not
         | generalize to web interfaces and more specifically, _how_ to do
         | those actions on those interfaces?
         | 
         | Not to take away from what this paper's scope and achievements.
        
       | 33a wrote:
       | How does this stack up to diffusion policy learning?
       | 
       | https://diffusion-policy.cs.columbia.edu/
       | 
       | Looking at the figures and videos it seems... worse? Sort of
       | surprised they didn't compare it but I guess they're trying to
       | limit the discussion to purely reinforcement learning methods.
       | 
       | EDIT: Ah, I see this is an older result and was published
       | concurrently to the diffusion policy paper, so it's likely the
       | authors didn't know about it in time to add the extra
       | comparisons.
        
       | nabakin wrote:
       | To save you guys some time, this is a 2 month old paper and not
       | made by OpenAI. It could be related to Q* (I don't know enough to
       | determine that) but it's not OpenAI's Q*.
        
       | osti wrote:
       | Is Q the new hottest letter right now?
        
         | smokel wrote:
         | To make this strand of discussion slightly more interesting, I
         | hereby present you a page [1] with the history of Q-tips.
         | Apparently, the "Q" stands for Quality :)
         | 
         | [1] https://www.qtips.com/about/
        
         | nothrowaways wrote:
         | Q is nothing without the *
        
           | GTP wrote:
           | You're the Q to my *!
        
         | Filligree wrote:
         | This does seem pretty similar to Q-learning. So yes, things-
         | based-on-Qs are hot right now.
        
       ___________________________________________________________________
       (page generated 2023-11-30 23:02 UTC)