Lightnews — Scholar-powered news

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · 22d

My talk at Samsung AI Forum yesterday
www.youtube.com/watch?v=L2nA...

LRMs and Agentic AI (Talk at Samsung AI Forum)

YouTube video by Subbarao Kambhampati

www.youtube.com

1

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · 24d

In the year since LRMs ("reasoning models") hit the scene, we have been trying to understand, analyze and demystify them.. Here are our efforts to date--conveniently all in one place..👇

www.linkedin.com/posts/subbar...

In the year since LRMs ("reasoning models") hit the scene, we have been trying to understand, analyze and demystify them.. Here are our efforts to date--conveniently all in one… | Subbarao K...

In the year since LRMs ("reasoning models") hit the scene, we have been trying to understand, analyze and demystify them.. Here are our efforts to date--conveniently all in one place.. (𝗙𝗶𝗿𝘀𝘁..) 𝗘𝘃𝗮𝗹...

www.linkedin.com

1 4

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · 28d

𝐏𝐞𝐫𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐯𝐞 𝐓𝐡𝐢𝐧𝐤𝐢𝐧𝐠? The anthropomorphization of LRM intermediate tokens as thinking begat a cottage industry to "get efficiency by shortening thinking." We ask: 𝗜𝘀 𝗖𝗼𝗧 𝗹𝗲𝗻𝗴𝘁𝗵 𝗿𝗲𝗮𝗹𝗹𝘆 𝗮 𝗿𝗲𝗳𝗹𝗲𝗰𝘁𝗶𝗼𝗻 𝗼𝗳 𝗽𝗿𝗼𝗯𝗹𝗲𝗺 𝗵𝗮𝗿𝗱𝗻𝗲𝘀𝘀 𝗼𝗿 𝗶𝘀 𝗶𝘁 𝗺𝗼𝗿𝗲 𝗽𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝘃𝗲? 👉 www.linkedin.com/posts/subbar...

4

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · Aug 31

Rejecting papers in #AI Conferences because of "resource constraints" is shooting ourselves in the foot as a community; use Findings.. #SundayHarangue 👇

x.com/rao2z/status...

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) on X: "Rejecting papers in #AI Conferences because of "resource constraints" is shooting ourselves in the foot as a community; use Findings.. #SundayHarangue By now, we have all know that top AI conferences are oversubscribed (in terms of paper submissions), and have heard that that" / X

Rejecting papers in #AI Conferences because of "resource constraints" is shooting ourselves in the foot as a community; use Findings.. #SundayHarangue By now, we have all know that top AI conferences are oversubscribed (in terms of paper submissions), and have heard that that

x.com

2 1

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · Jul 22

Proofs are not reasoning traces & I/O Format Language shouldn't be much of an issue for LLMs + other things #SundayHarangue (Special IMO edition). 🧵 👇

x.com/rao2z/status...

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) on X: "Proofs are not reasoning traces & I/O Format Language shouldn't be much of an issue for LLMs #SundayHarangue (Special IMO edition). 1/ My feed these last couple of days of IMO discussions has been full of comments that seem to conflate LRM intermediate tokens (aka reasoning" / X

Proofs are not reasoning traces & I/O Format Language shouldn't be much of an issue for LLMs #SundayHarangue (Special IMO edition). 1/ My feed these last couple of days of IMO discussions has been full of comments that seem to conflate LRM intermediate tokens (aka reasoning

x.com

1 4

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · Jul 19

Both LLMs and LRMs are upper bounded by humanity's knowledge closure. True scientific discoveries are, by definition, outside of that closure. Ergo, LLMs/LRMs are great force multipliers to us; but don't support "Nobel this weekend" hype..

👉 www.linkedin.com/posts/subbar...

Neither LLMs nor LRMs have the ability to go beyond the humanity's knowledge closure--which is needed for true discoveries. | Subbarao Kambhampati

Neither LLMs nor LRMs have the ability to go beyond the humanity's knowledge closure--which is needed for true discoveries. Both are beholden to the collected knowledge of the humanity (whether de...

www.linkedin.com

2 9

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · Jul 13

Computational Complexity is the wrong measure for LRMs (as it was for LLMs)--think distributional distance instead #SundayHarangue (yes, we're back!)

👉 x.com/rao2z/status...

2

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · Jun 23

A̶̶̶I̶̶̶ ̶ ̶ ̶ ̶(̶A̶r̶t̶i̶f̶i̶c̶i̶a̶l̶ ̶I̶n̶t̶e̶l̶l̶i̶g̶e̶n̶c̶e̶)̶
̶̶̶A̶̶̶G̶̶̶I̶̶̶ ̶(̶A̶r̶t̶i̶f̶i̶c̶i̶a̶l̶ ̶G̶e̶n̶e̶r̶a̶l̶ ̶I̶n̶t̶e̶l̶l̶i̶g̶e̶n̶c̶e̶)̶
̶̶̶A̶̶̶S̶̶̶I̶̶̶ ̶(̶A̶r̶t̶i̶f̶i̶c̶i̶a̶l̶ ̶S̶u̶p̶e̶r̶ ̶I̶n̶t̶e̶l̶l̶i̶g̶e̶n̶c̶e̶)
ASDI (Artificial Super Duper Intelligence)

Don't get stuck with yesterday's hypeonyms!
Dare to get to the next level!

#AIAphorisms

1 3

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · Jun 19

This series of lectures was given the same week there was all that brouhaha over the Apple illusion paper (I was giving these lectures during the day and talking to reporters in the evening 😅). As such they are pretty up-to-date! 3/

x.com/rao2z/status...

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) on X: "Some of what that recent Apple LRM limitations paper shows is known (pardon my friendly Schmidhubering; I do welcome more LLM studies with scientific skepticism). Our study 👇 from Sep 2024 shows o1 accuracy degrading as complexity increases.. 1/ https://t.co/d8zEUGi4SZ" / X

Some of what that recent Apple LRM limitations paper shows is known (pardon my friendly Schmidhubering; I do welcome more LLM studies with scientific skepticism). Our study 👇 from Sep 2024 shows o1 accuracy degrading as complexity increases.. 1/ https://t.co/d8zEUGi4SZ

x.com

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · Jun 19

The lectures start with a "big picture" overview (Lecture 1); focus on standard LLMs and their limitations, and LLM-Modulo as a test-time scaling approach (Lecture 2); and end with a critical appraisal of the test-time scaling and RL post-training techniques (Lecture 3). 2/

1

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · Jun 19

For anyone interested, here are the videos of the three ~50min each lectures on the reasoning/planning capabilities of LLMs/LRMs that I gave at #ACDL2025 in Riva Del Sole resort last week. 1/

www.youtube.com/playlist?lis...

ACDL Summer School Lectures on Planning/Reasoning Abilities of LLMs/LRMs - YouTube

www.youtube.com

1 2 3

Reposted by Subbarao Kambhampati (కంభంపాటి సుబ్బారావు)

Melanie Mitchell @melaniemitchell.bsky.social · Jun 9

...it basically confirmed what is already well-established: LLMs (& LRMs & "LLM agents") have trouble w/ problems that require many steps of reasoning/planning.

See, e.g., lots of recent papers by Subbarao Kambhampati's group at ASU. (2/2)

2 5 52

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · Jun 16

An AGI-wannabe reasoning model whining that it couldn't handle a problem because its context window isn't big enough is like a superman-wannabe little kid protesting that he couldn't add those numbers because he doesn't have enough fingers and toes.. #AIAphorisms

3

Reposted by Subbarao Kambhampati (కంభంపాటి సుబ్బారావు)

Dr Abeba Birhane @abeba.bsky.social · Jun 1

"our counter-intuitive results demonstrate ways in which common interpretations of Large Reasoning Models may be anthropomorphizations or simplifications" arxiv.org/abs/2505.13775

Beyond Semantics: The Unreasonable Effectiveness of Reasonless Intermediate Tokens

Recent impressive results from large reasoning models have been interpreted as a triumph of Chain of Thought (CoT), and especially of the process of training on CoTs sampled from base LLMs in order to...

arxiv.org

2 11 55

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · May 30

The transformer expressiveness results are often a bit of a red herring as there tends to be a huge gap between what can be expressed in transformers, and what can be learned with gradient descent. Mind the Gap, a new paper with
Lucas Saldyt dives deeper into this issue 👇👇

x.com/SaldytLucas/...

Lucas Saldyt on X: "Neural networks can express more than they learn, creating expressivity-trainability gaps. Our paper, “Mind The Gap,” shows neural networks best learn parallel algorithms, and analyzes gaps in faithfulness and effectiveness. @rao2z https://t.co/8YjxPkXFu0" / X

Neural networks can express more than they learn, creating expressivity-trainability gaps. Our paper, “Mind The Gap,” shows neural networks best learn parallel algorithms, and analyzes gaps in faithfulness and effectiveness. @rao2z https://t.co/8YjxPkXFu0

x.com

3

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · May 28

Anthropomorphization of intermediate tokens as reasoning/thinking traces isn't quite a harmless fad, and may be pushing LRM research into questionable directions.. So we decided to put together a more complete argument. Paper 👉 arxiv.org/pdf/2504.09762 (Twitter thread: x.com/rao2z/status...)

1 10

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · May 25

Longer thread here

x.com/rao2z/status...

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) on X: "This RLiNo? paper lead by @soumya_samineni & @durgesh_kalwar dives into the MDP model used in the RL post-training methods inspired by DeepSeek R1, and asks if some of the idiosyncrasies of RL in R1 aren't just consequences of the simplistic structural assumptions in the MDP 🧵1/ https://t.co/qPBY3tJILE" / X

This RLiNo? paper lead by @soumya_samineni & @durgesh_kalwar dives into the MDP model used in the RL post-training methods inspired by DeepSeek R1, and asks if some of the idiosyncrasies of RL in R1 aren't just consequences of the simplistic structural assumptions in the MDP 🧵1/ https://t.co/qPBY3tJILE

x.com

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · May 25

This RLiNo? paper (arxiv.org/abs/2505.13697) lead by Soumya Samineni and Durgesh_kalwar dives into the MDP model used in the RL post-training methods inspired by DeepSeek R1, and asks if some of the idiosyncrasies of RL aren't just consequences of the simplistic structural assumptions made

1 4

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · May 21

Do Intermediate Tokens Produced by LRMs (need to) have any semantics? Our new study 👇

Thread 👉 x.com/rao2z/status...

2 8

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · May 13

Delighted to share that Siddhant Bhambri & Mudit Verma's
critical evaluation and refutation of the reasoning claims of ReACT has been accepted to #TMLR (Transactions on Machine Learning)

👉https://openreview.net/forum?id=aFAMPSmNHR

1 1 4

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · May 12

Solving Single Agent Fully Observable Deterministic (SAFODP) Problems with Dec-POMDP approaches #SundayHarangue #allegory

x.com/rao2z/status...

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) on X: "Solving Single Agent Fully Observable Deterministic (SAFODP) Problems with Dec-POMDP approaches #SundayHarangue #allegory Imagine you modeled your decision problem into a Dec-POMDP problem ('cuz that's as expressive a decision model as you can get! )--but with some https://t.co/vDWYTHbnQA" / X

Solving Single Agent Fully Observable Deterministic (SAFODP) Problems with Dec-POMDP approaches #SundayHarangue #allegory Imagine you modeled your decision problem into a Dec-POMDP problem ('cuz that's as expressive a decision model as you can get! )--but with some https://t.co/vDWYTHbnQA

x.com

2

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · May 9

IMHO, the whole idea of connecting "length of intermediate tokens" produced by LRMs to inference time compute is a mind-boggling demonstration of circular reasoning--that comes from the assumptions about MDP model and reward model.. 👇

x.com/rao2z/status...

x.com

1 3

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · May 5

It ain't "The Bitter Lesson" if you are in the loop curating the training data for your LLM, y'all.. Pick your lesson, will ya? #SundayHarangue (h/t @kstechly.bsky.social)

3 4

Reposted by Subbarao Kambhampati (కంభంపాటి సుబ్బారావు)

Adrian Chan @gravity7.bsky.social · Apr 19

Don't use summarizers for the papers by @rao2z.bsky.social because the reasoning traces therein are, unlike the LRMs & LLMs under investigation, substantively meaningful, semantically well-ordered, and stylistically compelling and engaging!
#AI #LLMs #CoT
arxiv.org/abs/2504.09762

(How) Do reasoning models reason?

We will provide a broad unifying perspective on the recent breed of Large Reasoning Models (LRMs) such as OpenAI o1 and DeepSeek R1, including their promise, sources of power, misconceptions and limit...

arxiv.org

1 2 8

Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z.bsky.social · Apr 16

Here is a recording of my talk at @msftresearch.bsky.social last week titled "(How) Do LLMs Reason/Plan?" (Also gave a version of it at as a distinguished lecture at Oracle today..)

www.youtube.com/watch?v=0u2h...

(How) Do LLMs Reason/Plan? (Talk given at Microsoft Research; 4/11/25)

YouTube video by Subbarao Kambhampati

www.youtube.com

1 5