[HN Gopher] NeurIPS 2025 Best Paper Awards
       ___________________________________________________________________
        
       NeurIPS 2025 Best Paper Awards
        
       Author : ivansavz
       Score  : 162 points
       Date   : 2025-12-05 01:15 UTC (21 hours ago)
        
 (HTM) web link (blog.neurips.cc)
 (TXT) w3m dump (blog.neurips.cc)
        
       | Scene_Cast2 wrote:
       | I think my favorite of the bunch is the "Does Reinforcement
       | Learning Really Incentivize Reasoning Capacity in LLMs Beyond the
       | Base Model" paper. Easy to read, gets the point across very
       | intuitively and quickly, and the point is very interesting and
       | relevant to a lot of people.
       | 
       | About the Superposition paper - this is close to what I've been
       | thinking about over the past week. I'm thinking that concepts or
       | choices in a "superposition" are harder for a fully-
       | differentiable neural net to reason about. For example, if
       | there's a "green" vs "purple" choice to be made, it can't fully
       | commit to either (especially if they're 50-50), and will have to
       | reason about both simultaneously (difficult due to nonlinear
       | manifold space). Discretizing to tokens (non-differentiable
       | argmax) forces a choice, and that allows it to reason about a
       | single concept separately and easier.
        
         | energy123 wrote:
         | I am not sure how to interpret the first paper's results.
         | 
         | If we use a random number generator then we will converge to
         | 100% correct answers under pass@n in the limit.
         | 
         | A random number generator will eventually outperform or match
         | all models (for large n) whenever top-p is less than 1 because
         | the other models will most likely have some level of bias that
         | makes correct CoTs mathematically impossible due to the tokens
         | being too improbable and being filtered out by top-p, meaning
         | that other models will asymptote to below 100% while the RNG
         | will reach 100% in an almost surely sense.
         | 
         | Under this paper's logic doesn't that mean that the random
         | number generator is a superior reasoner?
        
           | Scene_Cast2 wrote:
           | I'm not sure how likely it is that an answer would fall
           | outside of the top-p of 0.95 (used in the paper). A random
           | number generator would also need an unreasonably high number
           | of samples to get a correct answer. I think figures 17 and 18
           | are interesting for this discussion too, they show
           | performance at various sampling temperatures. I think the
           | point of the paper is that RL "sharpens" the distribution of
           | non-RL nets, but it does not uncover any new reasoning paths
           | - non-RL nets already had multiple decently high probability
           | paths of answering questions to begin with, and RL reuses a
           | subset of those.
        
           | robrenaud wrote:
           | I agree that pass@k feels a bit weird for large k. But for
           | LLMs, it's a decent proxy for "are the
           | knowledge/skills/circuit necessary to solve the problem
           | somewhere in the model". Note that choices for large k is on
           | the order of 256, and the range of valid answers is much
           | larger than that. So your infinite monkeys critique, while
           | true in the limit, wouldn't actually outperform models in the
           | tested regime.
           | 
           | Also, in practice, models don't have that much semantic
           | entropy of a given prompt. With temperature based sampling,
           | models will tend to generate very similar but not identical
           | responses.
        
           | boroboro4 wrote:
           | To me intellect has two parts to it: "creativity" and
           | "correctness". And from this perspective random sampler is
           | infinitely "creative" - over (infinite) time it can come up
           | with answer to any given problem. And from this perspective
           | it does feel natural that base models are more "creative"
           | (because that's what being measured in the paper), while RL
           | models are more "correct" (that's a slope of the curve from
           | the paper).
        
       | ilaksh wrote:
       | Does some have a similar award for papers that are innovative?
       | Like new, relatively unproven architectures?
        
         | robrenaud wrote:
         | From what I've seen at neurips, in terms of most different but
         | maybe viable, it would be this.
         | 
         | https://sakana.ai/ctm/
         | 
         | In terms of a fresh perspective on designing learning systems,
         | nested learning seems very interesting.
         | 
         | https://abehrouz.github.io/files/NL.pdf
         | 
         | Hearing the clarity, creativity, and force behind his thoughts
         | and speech, I'd give a more than 1/200 chance Ali Behrouz gets
         | himself a Turing award. At the very least, I think he will end
         | making major contributions to AI.
        
         | djrhails wrote:
         | There is TITANS - https://arxiv.org/abs/2501.00663
        
       | chermi wrote:
       | Interesting that 3 names I recognized as physicists from stat
       | mech adjacent fields. They continue to punch above their
       | expectations (as sampled by general dismissal of physicists in
       | AI/ML on HN and reddit).
        
         | chatmasta wrote:
         | Some of the best software engineers I know are ex-physics
         | PhDs... it's one of those "can't fake it" skillsets that also
         | happens to have high transferability to ML/AI fields. On the
         | other hand, I snuck through the CS major without ever
         | multiplying a matrix.
        
           | ctxc wrote:
           | Haha, nice bio. Seeing that font on HN is quite a shock.
        
           | miki123211 wrote:
           | > I snuck through the CS major without ever multiplying a
           | matrix
           | 
           | I didn't, but only because I became personally interested in
           | AI/ML at some point, so I actually had to learn it myself.
           | 
           | As an AI practitioner, I still couldn't explain eigenvectors
           | or singular-value decomposition to you though.
        
         | mnky9800n wrote:
         | Do people not like physicists?
        
           | jmalicki wrote:
           | https://xkcd.com/793/ captures the stereotype well.
        
             | peterfirefly wrote:
             | Especially because those annoying dilettante know-it-all
             | physicists are often right.
        
       | niceguy4 wrote:
       | Are there any talks about these papers on youtube or somewhere? I
       | think I find it easier to listen and watch then read or maybe I'm
       | just lazy, not sure.
        
         | neves wrote:
         | There conference had interesting lectures. Will they be posted
         | online?
        
         | cosmic_ape wrote:
         | most papers have slides with audio, and some, including the
         | awards ones will have short frontal talks. this will be
         | released at some point after the conference, but right now
         | looks like you'd have to be registered to see it.
        
         | FrozenSynapse wrote:
         | use NotebookLM
        
           | cubefox wrote:
           | Whenever I search for the title a new machine learning paper,
           | there are a bunch of YouTube videos about it which are just
           | NotebookLM slop. It's straight up environmental pollution.
        
             | Der_Einzige wrote:
             | One of the most popular of those slop videos was about our
             | antislop sampler. Ironic.
             | 
             | https://youtu.be/PHSqcdIc5gM?si=I62bduoDgnlNFPZ6
        
       | gradascent wrote:
       | From the figure in the first paper listed:
       | 
       | > Responses to the query "Write a metaphor about time" clustered
       | by applying PCA to reduce sentence embeddings to two dimensions.
       | [...] The responses form just two primary clusters: a dominant
       | cluster on the left centered on the metaphor "time is a river,"
       | and a smaller cluster on the right revolving around variations of
       | "time is a weaver."
       | 
       | I just gave Gemini 3 the same prompt and got something quite
       | different:
       | 
       | >Time is a patient wind against the cliff face of memory. It does
       | not strike with a hammer to break us; it simply breathes, grain
       | by grain, until the sharp edges of grief are smoothed into
       | rolling hills, and the names we thought were carved in stone are
       | weathered into soft whispers.
        
         | SiempreViernes wrote:
         | Constantly flowing and makes things smooth like river stones;
         | compared to Tait's "time is a series if staric pictures",
         | Gemini's output is not so different from a river metaphor.
        
       | djoldman wrote:
       | Oh man, this link is worth it just for the "Reflections from the
       | Selection Committee."
       | 
       | These days, abstracts are so marketing/advertising forward that
       | it's hard to even understand the claim.
        
       | yanhangyhy wrote:
       | seems lots of chinese.
        
       ___________________________________________________________________
       (page generated 2025-12-05 23:01 UTC)