Rose @rose_e_wang Twitter profile

Pinned Tweet

Rose

7 days

What do great teachers do to be good at teaching? What can this teach us about LMs? Most of us experience the “front stage” of teaching—as students. Few see the *back stage*: the planning, pedagogical decisions… 🌉 Bridge, at NAACL’24, surfaces these hidden decisions🧵

3

16

65

Last Seen Profiles

@FreshKnight1

@hilltommi13

@MaameBlueWrites

@tsprada790312

@stw_pdg

@MatrixedLink

@AvDomo19

@Kellzallday

@Havas

@airimania_staff

@_gregomerchann_

@MariaVazqu64922

@unrealhybrid

@mikeclindsay

@JohnnySinn99691

@corn_pirate

@BlinkJQ

@charleau0029

@Juancarlos46387

@skyyina

@xopyc_akcimand

@JanaBeris1

@xiaokucat88617

@ilyfuego

@_lili_i

@MesousaGaby

@Lilieth6969

@AdalidMendozaF1

@AlexanderA43698

@Jay_Stevenson7

@LensaHijau09

@CodecIreland

@KellySlayAnn

@crewgrrl

@dschirwon

@VanillaADC

Rose

@rose_e_wang

2 years

I’m excited to share “Language modeling via stochastic processes”, an #ICLR2022 oral! Our work addresses the challenge of generating long _coherent_ sequences with language models by leveraging goal-conditioned latent paths. Paper: 🧵[1/T]

8

110

663

Rose

@rose_e_wang

4 years

Excited to share gym-cooking, a *novel multi-agent Gym environment*: Based on recent work (, #CogSci2020 computational modeling prize winner) with amazing collaborators Sarah Wu, James Evans, Josh Tenenbaum, David Parkes, @maxkw !

6

60

305

Rose

@rose_e_wang

2 years

How can we algorithmically figure out what our model doesn’t know, and then construct datasets to improve it? We tackle this question in “Know thy student: Interactive learning with Gaussian processes” at #ICLR2022 @cells2societies workshop. Paper: [1/N]

1

23

192

Rose

@rose_e_wang

2 months

Students listening to lectures can go to Google to answer their questions. But…how can the teacher find what part of their lecture *caused* the student’s question in the first place?? 🤔 Introducing *Backtracing*: Retrieving the Cause of the Query! 🧵

8

37

188

Rose

@rose_e_wang

4 months

There’s a *lot* of insight about how students learn in education data like classroom conversations, but I can tell you it's *painful* to process and analyze. Introducing *Edu-ConvoKit*, a pipeline that handles the pre-processing, annotation, and analysis for you! 🧵

2

19

95

Rose

@rose_e_wang

3 years

How do we train language models (LMs) to be good pragmatic conversational partners? We investigate this in our #EMNLP2021 Findings paper: Calibrate your listeners! Robust communication-based training for pragmatic speakers. 📜: 📺:

EMNLP 2021: Calibrate your listeners! Robust communication-based...

Presented at Findings of EMNLP 2021Calibrate your listeners! Robust communication-based training for pragmatic speakersRose E. Wang, Julia White, Jesse Mu, N...

www.youtube.com

2

9

60

Rose

@rose_e_wang

2 months

#EACL2024 is over!! This was a super fun conference & it was great meeting new folks 😃 Malta is so beautifully colorful --- here are some of my favorite pics!! ✨🎨

2

57

Rose

@rose_e_wang

7 months

Ever wonder how experienced math teachers & tutors compare to ChatGPT or GPT4 in teaching students? 🖥️🧑‍🎓👩‍🏫 Check out our new paper “Step-by-Step Remediation of Students’ Mathematical Mistakes”! 📜 🖥️ from @stanfordnlp @StanfordEd

2

22

51

Rose

@rose_e_wang

10 months

Can ChatGPT help teachers by providing effective feedback, like generating helpful pedagogical suggestions? 👩‍🏫 We answer this question in our work presented @ BEA (co-hosted @ ACL) on Thursday July 13 Harbour A! Website: w/ @ddemszky !

1

12

47

Rose

@rose_e_wang

2 years

Come check out “Language modeling via stochastic processes” at #ICLR2022 on this Monday! 😄 Looking forward to meeting old and new friends 🥳 Livestream: Apr 25 5-5:15pm PT Poster: Apr 25 6:30-8:30pm PT w/ @esindurmusnlp , Noah Goodman & @tatsu_hashimoto

Rose

@rose_e_wang

2 years

I’m excited to share “Language modeling via stochastic processes”, an #ICLR2022 oral! Our work addresses the challenge of generating long _coherent_ sequences with language models by leveraging goal-conditioned latent paths. Paper: 🧵[1/T]

8

110

663

0

7

44

Rose

@rose_e_wang

3 years

It's finally time! Join us tomorrow (Friday) for a series of exciting discussions on the roadmap to never-ending RL with the amazing @danijarh @aharutyu @joelbot3000 @pyoudeyer @NataliaDiazRodr @AleksandraFaust @hyogweon and more! 🥳🥳🥳 Schedule:

NERL2021

Motivation

sites.google.com

Never Ending RL

@neverendingrl

3 years

1/N We are excited to introduce our @iclr_conf workshop: A Roadmap to Never-Ending RL. We invite you to submit papers (up to 6 pages, excluding references and appendix) in the @iclr_conf format. Submission Deadline: February 26, 2021 #NERL2021 #ICLR2021

1

29

112

1

9

37

Rose

@rose_e_wang

3 years

@Shreyagupta08 I’M HIS PHD STUDENT AND I’VE NEVER SEEN HIM IN PERSON I’M SCREAMINGGGGG

2

0

36

Rose

@rose_e_wang

2 months

In Malta 🇲🇹 for EACL this week!! Let me know if you’re around and wanna talk about nlp applications esp education, or just hang with me @chengmyra1 @krisgligoric … or run with me along the coast!! 😎🏃🏻‍♀️

0

4

31

Rose

@rose_e_wang

2 months

Presenting Backtracing in Malta 🇲🇹 tomorrow at EACL 14:00 in Session A, Radisson Ballroom Level 2 — come say hi!!! 😃

Rose

@rose_e_wang

2 months

Students listening to lectures can go to Google to answer their questions. But…how can the teacher find what part of their lecture *caused* the student’s question in the first place?? 🤔 Introducing *Backtracing*: Retrieving the Cause of the Query! 🧵

8

37

188

0

4

29

Rose

@rose_e_wang

22 days

How will ed tech change w LLMs? What is and isn't possible? If these Qs have been on your mind, submit your work to a workshop I'm organizing: Leveraging LLMs for Next Gen Ed Tech @ EDM 2024 by May 10th! ➡️ #EDM #EdTech

EDM Workshop

Description

sites.google.com

2

5

28

Rose

@rose_e_wang

3 years

Curious about how we can build cooperative, human-like AI systems? 🤖 📜: Poster 10-11am PST Saturday #NeurIPS2020 Cooperative AI Workshop! 🗣: Spotlight talk 11:45pm PT Saturday 📚: Paper w/ S. Wu, J. Evans, J. Tenenbaum, D. Parkes, @maxhkw !

Too many cooks: Bayesian inference for coordinating multi-agent...

Collaboration requires agents to coordinate their behavior on the fly, sometimes cooperating to solve a single task together and other times dividing it up into sub-tasks to work on in parallel....

arxiv.org

1

2

26

Rose

@rose_e_wang

3 years

Woo, my internship project is out on the @GoogleAI blog!!! How can we get real-world robots to anticipate each other’s behavior and collaborate? Find out more below!👇 Big thanks to: J. Chase Kew, Dennis Lee, Tsang-Wei Lee, Tingnan Zhang, @brian_ichter , Jie Tan, @AleksandraFaust

Google AI

@GoogleAI

3 years

Introducing a model-based #RL approach for robot navigation, called hierarchical predictive planning (HPP), that enables agents to align their goals on the fly in order to solve the decentralized rendezvous task. Learn more at

7

197

818

0

1

26

Rose

@rose_e_wang

3 years

Excited to present #CoRL2020 paper on model-based RL for multirobot coordination! Work done at #Google w J. Chase Kew, Dennis Lee, Tsang-Wei Lee, Tingnan Zhang, @brian_ichter , Jie Tan, @AleksandraFaust .😄 Website: Live session: Tomorrow (11/18) 12:30pmPT

Hierarchical Predictive Planner

Links: [paper] [CoRL 5-minute talk][robot experiments]

sites.google.com

0

2

25

Rose

@rose_e_wang

8 days

📢 Calling the #EdTech community! Intrigued by the [potential/positive/negative] impact of LLMs on education? Submit your work to this workshop at #EDM2024 🪇👩‍🏫 ➡️ Deadline: May 10th Looking forward to the discussions!!!

EDM Workshop

Description

sites.google.com

Rose

@rose_e_wang

22 days

How will ed tech change w LLMs? What is and isn't possible? If these Qs have been on your mind, submit your work to a workshop I'm organizing: Leveraging LLMs for Next Gen Ed Tech @ EDM 2024 by May 10th! ➡️ #EDM #EdTech

2

5

28

3

6

26

Rose

@rose_e_wang

8 months

I am thrilled to present at Women in Data Science! Come learn more about my work with @ddemszky on NLP and Education 😄 Topic: "Beyond Right or Wrong: Leveraging Language Models to Enhance the Learning Process" Aug. 30, 11am-12pm PDT Register now:

1

2

25

Rose

@rose_e_wang

1 year

@ccanonne_ This is neat! Curious about what students wrote. One thing that’s been on my mind is how this might change grading rubrics to be less about a student’s generative ability (eg pseudocode, complexity analysis), and more about their discriminative ability (eg debug, critique).

1

20

Rose

@rose_e_wang

22 days

📢 Calling all folks in the AI x Education community: An exciting opportunity!!! 👩‍🏫👩‍🎓 Bill and Melinda Gates Foundation RFI on AI-Powered Innovations in Math Teaching & Learning:

AI-Powered Innovations in Mathematics Teaching & Learning Request for Information

One of the most powerful ways for students to be able to take charge of their own future is through success in math. Math teaches students to make sense of the world and how to be better problem...

usprogram.gatesfoundation.org

0

2

19

Rose

@rose_e_wang

2 months

Let's build a future where every question leads to deeper understanding…for both students and teachers! 💡 Big thanks to @arankomatsuzaki for featuring our Backtracing work!! 🙏

Aran Komatsuzaki

@arankomatsuzaki

2 months

Backtracing: Retrieving the Cause of the Query - Proposes a new task called backtracing where the goal is to retrieve the cause of the query from a corpus - Shows limitations in current retrieval methods for performing backtracing repo: abs:…

3

45

227

2

3

19

Rose

@rose_e_wang

2 years

Hi friends!! I'll be back at MIT for re-graduation on Monday for a week!! Let me know if you want to meet 🙏 + hang 🥳 + grab coffee ☕️ + eat food 🍜!! 😄

1

0

18

Rose

@rose_e_wang

2 years

This was done in collaboration with the amazing @esindurmusnlp , Noah Goodman and @tatsu_hashimoto ! 🥳 Many many thanks to my labmates from CoCoLab for their incredible support and feedback: @mike_h_wu , @GabrielPoesia , Ali Malik, @AlexTamkin ! 😎 🧵[T/T]

0

2

17

Rose

@rose_e_wang

2 years

Our insight is to represent _coherent_ language as a _smooth_ latent trajectory. We turn to Brownian bridge stochastic processes (SP) as a model for smooth trajectories. [4/T]

1

16

Rose

@rose_e_wang

2 years

Take a look at the paper/code/video for more details, other results and ablations! Paper: Code: Video: [10/T]

ICLR 2022 Oral: Language modeling via stochastic processes

Language modeling via stochastic processesRose E. Wang, Esin Durmus, Noah Goodman, Tatsu HashimotoICLR 2022 OralPaper: https://openreview.net/forum?id=pMQwKL...

www.youtube.com

2

0

17

Rose

@rose_e_wang

2 years

We train our encoder with contrastive learning (CL). Why CL? Bc of (a) its striking performance in learning representations (SimCLR for images @tingchenai ) & (b) exciting work in applying it to structured SPs ( @BingbinL )! [5/T]

2

0

15

Rose

@rose_e_wang

2 months

Understanding why users ask questions is key because it’s a natural source of *content feedback*. We establish a diverse benchmark for backtracing and show traditional retrieval systems miss the mark in retrieving the cause of queries 😬 📎:

1

0

14

Rose

@rose_e_wang

2 years

When we want language models to generate long text, they often output meandering text. One potential reason behind this failure mode is the model’s inability to plan ahead or represent long-range text dynamics. [2/T]

1

0

13

Rose

@rose_e_wang

4 months

You want to understand how students learn, but not sure where to get started? 🎥 Here's a 2-min demo video of Edu-ConvoKit! It walks through its GPT-powered, quantitative and qualitative analysis tools. 👋 Happy exploring with Edu-ConvoKit!

Edu-ConvoKit Demo Video

Welcome to Edu-ConvoKit! There’s a lot of insight about how students learn in education conversations, but it’s painful to process and analyze. Edu-ConvoKit ...

www.youtube.com

Rose

@rose_e_wang

4 months

There’s a *lot* of insight about how students learn in education data like classroom conversations, but I can tell you it's *painful* to process and analyze. Introducing *Edu-ConvoKit*, a pipeline that handles the pre-processing, annotation, and analysis for you! 🧵

2

19

95

0

3

12

Rose

@rose_e_wang

4 years

Our environment is designed to be easily configurable and light-weight. It’s perfect for folks interested in multi-agent systems or in compositional tasks/environments!

1

0

13

Rose

@rose_e_wang

10 months

Lectures are a learning experience for students & teachers. Students learn about the subject & teachers learn about refining their instruction. But, online student feedback is unstructured. How can teachers learn from it?

1

2

13

Rose

@rose_e_wang

2 months

Have questions about the work? Let's trace back to the source...: 📎: 💻: I’ll also be in Malta for #EACL presenting this work, so come chat with me there too 😉☀️

GitHub - rosewang2008/backtracing: Backtracing: Retrieving the Cause of the Query, EACL 2024 Long...

Backtracing: Retrieving the Cause of the Query, EACL 2024 Long Paper, Findings. - rosewang2008/backtracing

github.com

1

12

Rose

@rose_e_wang

2 years

On long text generation settings, Time Control (TC) preserves the text structure both in terms of ordering (up to +40% better) and text length consistency (up to +17% better). Human evaluators also prefer TC's output 28.6% more than the baselines. [8/T]

1

0

11

Rose

@rose_e_wang

2 years

Prior work has explored remedies for this failure mode by using planning-based methods or implicitly learning text dynamics. However, these methods manually specify the text dynamics or sacrifice quality in long-horizon generation. [3/T]

1

0

10

Rose

@rose_e_wang

3 years

Yes!!! Make sure to check out @jeffclune + @ruiwang2uiuc 's awesome work at our 2nd poster session at 12:55pm PT! 🥳 GatherTown link can be found on our workshop's ICLR site.

Jeff Clune

@jeffclune

3 years

Our work Enhanced POET is an invited poster at this (excellent!) workshop in case you want to come ask questions or ask about any of our work. Thanks to the organizers for putting together such a wonderful event!

3

6

43

0

4

11

Rose

@rose_e_wang

4 months

👉 Check Edu-ConvoKit out: It’s easy and instrumental to transform the way we conduct research for improving real student learning outcomes. Work done at @StanfordEd @stanfordnlp @StanfordAILab with my advisor @ddemszky

GitHub - stanfordnlp/edu-convokit: Edu-ConvoKit: An Open-Source Framework for Education Conversat...

Edu-ConvoKit: An Open-Source Framework for Education Conversation Data - stanfordnlp/edu-convokit

github.com

1

3

11

Rose

@rose_e_wang

7 days

Experienced teachers engage their students in critical thinking—whereas novice tutors and LLMs don’t: They frequently give away the answer. Our work focuses on how experienced teachers do and *think* about remediating student mistakes. 📎:

1

0

11

Rose

@rose_e_wang

2 years

The intuition is simple: The bridge imposes that a positive triplet (eg. three in-order sentences on Boston) makes up a smooth trajectory. A negative triplet should not construct a smooth trajectory (switching middle sentences with one on New York). [6/T]

1

0

10

Rose

@rose_e_wang

2 years

After training the encoder, we finetune GPT2 to decode from past context and the encoded latent plan. At inference, we generate a latent plan by sampling from the bridge and conditionally generate each sentence using the latent plan. [7/T]

1

0

10

Rose

@rose_e_wang

2 years

Lately, I’ve been super excited about teacher-student settings & thinking about how we can enable machines to (one day) reliably interact & _teach_ humans! If you’re interested in this direction, let's chat at @cells2societies poster session Fri April 29 8:15-9:05am PT! [6/N]

1

10

Rose

@rose_e_wang

10 months

(belated) paper laptop stickers for #ACL2023 !!! made with #stablediffusion 😃

1

0

9

Rose

@rose_e_wang

3 years

Join us at the @ResistanceAI NeurIPS workshop! Currently we have a fireside chat with the amazing @red_abebe , @timnitGebru , @Abebab , @tetisheri . @ResistanceAI aims to create an inclusive space, discuss critiques of how #AI currently concentrates power & formulate resistance tactics.

Resistance AI

@ResistanceAI

3 years

Our first panel of the day is starting shortly, at 9am PST! 💜

3

5

20

0

9

Rose

@rose_e_wang

2 years

Paper: w/ the amazing @mike_h_wu & Noah Goodman I’m grateful to many for sharing their insight, including my co-authors, Xiaojin Zhu (through his machine teaching works), @jayelmnop @DorsaSadigh @EmmaBrunskill @andyshih_ @willieneis @polynoamial [N/N]

0

8

Rose

@rose_e_wang

1 year

Super cool work from @KaitlynZhou --- generating expressions of uncertainty is extremely important in supporting human decision-making (and human reasoning)! My favorite part of the paper is their typology of uncertainty expressions (Table 4) 😀

Kaitlyn Zhou

@KaitlynZhou

1 year

Thanks so much for sharing our work! Our paper also discusses additional risks and opportunities that come with integrating expressions of uncertainty into LMs. Read the paper here: w/ @jurafsky @tatsu_hashimoto @stanfordnlp

0

2

15

0

4

8

Rose

@rose_e_wang

2 years

We also show that TC doesn’t sacrifice short/mid-range language modeling performance! Eg. TC matches/outperforms task-specific models like infilling by language modeling (ILM) on text-infilling or local representation methods on discourse coherence. [9/T]

1

0

8

Rose

@rose_e_wang

4 months

I am excited about language tools for education at scale because we move away from an oversimplified view of learning measured by the standardized test score….and towards language measures of student thinking and pedagogy. Repo:

1

8

Rose

@rose_e_wang

2 years

Stop by @cells2societies poster session tomorrow (Fri) 8:15-9:05am PT if you're interested in teacher-student settings or the problem of "assessing then teaching models". Always excited to meet new folks too! 🙏🙂

Rose

@rose_e_wang

2 years

How can we algorithmically figure out what our model doesn’t know, and then construct datasets to improve it? We tackle this question in “Know thy student: Interactive learning with Gaussian processes” at #ICLR2022 @cells2societies workshop. Paper: [1/N]

1

23

192

0

3

8

Rose

@rose_e_wang

2 months

What makes backtracing so hard?! Let’s use our Lecture dataset based on MIT OCW lectures as an example. Challenge 1: The queries don’t explicitly label what they’re caused by in the lecture/source document

1

0

8

Rose

@rose_e_wang

1 year

@janleike @BlancheMinerva @MaksimSTW It would be great if OpenAI could make a more public blog post (eg. something with as much visibility as ChatGPT on ), clarifying the misunderstandings and also the implications (eg. SL fine-tuning on a good dataset ~= RLHF, and not RLHF>SL)!

2

0

7

Rose

@rose_e_wang

4 months

Tremendous thanks goes to the @lateinteraction , @JimMalamut , @lucy3_li for their feedback---esp. @lateinteraction for all the amazing, punny name suggestions (unfortunately, they didn't make the cut in my Slack polls!)🥲 Thanks for @YannHicke for contributions to the toolkit!

1

0

7

Rose

@rose_e_wang

3 years

I’ll be presenting our work in person at EMNLP. This will be my first NLP & in-person conference, so I’d love to meet folks! Please don’t hesitate to reach out 😄 📜: 📺: 👩‍💻:

GitHub - rosewang2008/calibrate_your_listeners: Calibrate your listeners! Robust communication-ba...

Calibrate your listeners! Robust communication-based training for pragmatic speakers. Findings of EMNLP 2021. - rosewang2008/calibrate_your_listeners

github.com

0

7

Rose

@rose_e_wang

2 months

Challenge 2: The source document is *super* long. Long text understanding is important for backtracing to find the cause, but hard to nail.

1

0

7

Rose

@rose_e_wang

2 months

Come hang out with us if you’re at #EACL2024 !!!

Kristina Gligorić

@krisgligoric

2 months

Attending #EACL2024 and passionate about computational social science (CSS) or social applications of NLP? Join us tomorrow (Wednesday at 2 pm in the Carlson room) for an informal CSS Birds of a Feather gathering! Let's meet, chat, and share insights!

2

14

47

0

7

Rose

@rose_e_wang

3 years

Co-organized with the great @FeryalMP , @khimya , @LouisKirschAI , Annie Xie, Adam White, and Doina Precup at @iclr_conf --- see you all then!!!

0

7

Rose

@rose_e_wang

2 months

And Challenge 3: Backtracing relies domain expertise to understand the source document, the queries and the *reasons* behind the student’s confusion.

1

0

7

Rose

@rose_e_wang

7 months

I'm excited about building NLP algs/systems to empower educators & students and enhance social interactions! ✨🚀🧑‍🎓👩‍🏫 Echoing the views of @ddmeyer in education and the innovative NLP x data science approaches of @Diyi_Yang & @timalthoff in mental health

1

0

6

Rose

@rose_e_wang

7 months

@abeirami @emnlpmeeting When can we expect a response from the PCs? We sent an email on Sunday and still haven't heard back. Our soundness/excitement scores were 5/4, 4/4, 4/4... 🙏 @juanmiguelpino @hbouamor Totally understand it's a busy time, but having an ETA would be helpful for resubmission! 🙂

2

0

5

Rose

@rose_e_wang

2 months

And thank you to @_akhaliq for his thread on our work as well 😃

AK

@_akhaliq

2 months

Backtracing Retrieving the Cause of the Query Many online content portals allow users to ask questions to supplement their understanding (e.g., of lectures). While information retrieval (IR) systems may provide answers for such user queries, they do not directly assist

2

16

110

1

0

6

Rose

@rose_e_wang

2 months

And finally, a huge shoutout to my collaborators from @stanfordnlp @StanfordEd , esp @lateinteraction ! Working with Omar has been an incredible experience: He brings boundless insight and energy to the table, and I’m so grateful to learn from this star in our community! 🌟

2

0

6

Rose

@rose_e_wang

7 days

We use Cognitive Task Analysis & work with experienced 🧑‍🏫s to surface their internal decisions. Patterns emerge: 🧑‍🏫 infer the *error type* -> determine an *intention* -> pick a *remediation strategy*.

1

0

7

Rose

@rose_e_wang

7 months

We have a lot of other exciting ongoing projects along this direction, so stay tuned!!! 📜 👩‍💻 🖥️ w/ the amazing Ashley Zhang, Carly Robinson, Susanna Loeb, and @ddemszky ✨

Step-by-Step Remediation of Students' Mathematical Mistakes

Rose E. Wang, Qingyang Zhang, Carly Robinson, Susanna Loeb, Dorottya Demszky. Step-by-Step Remediation of Students' Mathematical Mistakes. 2020.

rosewang2008.github.io

2

6

Rose

@rose_e_wang

7 days

Our work complements recent work on self-improving LLMs with internal decision-making: But, we focus on eliciting and leveraging thoughts/decisions from real, experienced humans—in this case, teachers!

Eric Zelikman ✈️ ICLR

@ericzelikman

2 months

Language models today are trained to reason either 1) generally, imitating online reasoning data or 2) narrowly, self-teaching on their own solutions to specific tasks Can LMs teach themselves to reason generally?🌟Introducing Quiet-STaR, self-teaching via internal monologue!🧵

35

189

978

1

0

6

Rose

@rose_e_wang

7 days

We contribute a *unique, real* dataset: It contains the *internal decisions* paired with the response of these experienced teachers. It also includes real tutoring conversation snippets between novice tutors and students, across 120 math topics. Dataset:

1

0

6

Rose

@rose_e_wang

7 days

There are so many cool findings in this work—I could go on for ages, but tweets can only do so much… For now I’ll leave y’all on a cliffhanger and show how experienced teachers make *diverse* and *complex* decision paths, compared to LLMs. Isn’t this beautiful!?

2

0

7

Rose

@rose_e_wang

2 months

@ChengleiSi @StevenyzZhang @zhengyuan_yang @Microsoft @RuiboLiu @GoogleDeepMind @Diyi_Yang @stanfordnlp @shi_weiyan @gaotianyu1350 @WilliamBarrHeld Next time I'd like to contribute to the good vibes too!! 😃😆 Big congrats @ChengleiSi & @StevenyzZhang on this work!!!

1

0

5

Rose

@rose_e_wang

2 years

Eg. in an offline reinforcement learning setting, the student must navigate to the goal (green). The teacher determines states (yellow) the student has explored and accomplishes this task. The teacher can then construct demonstrations from states (orange) the student fails. [3/N]

1

0

5

Rose

@rose_e_wang

3 years

Get ready for our panel happening in 10 mins!!! with @celestekidd @MelMitchell1 @SchmidhuberAI Adam White and Satinder Singh!!! 😀

Never Ending RL

@neverendingrl

3 years

Please join us for a roundtable panel discussion with Celeste Kidd ( @celestekidd ), Satinder Singh, Melanie Mitchell ( @MelMitchell1 ), and Jürgen Schmidhuber ( @SchmidhuberAI ) moderated by Adam White. We are excited to hear this discussion. Don’t miss it!

0

2

12

0

1

5

Rose

@rose_e_wang

4 months

@aryaman2020 @ChengleiSi Meanwhile @chengmyra1 , @krisgligoric and me *also* in the office 👀👀👀

1

0

5

Rose

@rose_e_wang

2 years

Make sure to check out @dharakyu & @jayelmnop 's cool work tomorrow (Friday) @EmeComWorkshop #ICLR2022 ! 🥳

Dhara Yu

@dharakyu

2 years

Excited to present “Emergent Covert Signaling in Adversarial Reference Games” at #ICLR2022 @EmeComWorkshop ! Paper: A thread: (1/n)

1

7

18

0

4

Rose

@rose_e_wang

1 year

tfw you pay for chatgpt+ but only get text-davinci-002 😢

2

0

3

Rose

@rose_e_wang

3 years

Looking for a library AND viz tool for making ML training more efficient? Want to understand the trade offs you would make between cost, time and model quality? Check out what my cool friends @AveryLamp @abhi_venigalla @moinnadeem at @mosaicML recently released!!! 🥳😎

Databricks Mosaic Research

@DbrxMosaicAI

3 years

Hello World! Today we come out of stealth to make ML training more efficient with a mosaic of methods that modify training to improve speed, reduce cost, and boost quality. Read our founders' blog by @NaveenGRao @hanlintang @mcarbin @jefrankle (1/4)

7

41

164

0

4

Rose

@rose_e_wang

28 days

@aryaman2020 @ChengleiSi This is so wholesome ❤️

0

4

Rose

@rose_e_wang

7 days

This work couldn’t have happened without the amazing support of my interdisciplinary team (CS+Education+Economics), and teachers of course! @stanfordnlp @StanfordEd @StanfordAILab Ashley @Carly__Robinson @loeb_susanna @ddemszky

1

0

5

Rose

@rose_e_wang

7 days

Our Bridge framework allows for two cool questions to be answered: - Can we use a human expert’s internal decisions to *improve* LLM responses? - Can we prompt LLMs to make their own internal decisions and self-improve?

1

0

5

Rose

@rose_e_wang

7 days

And yes, we *can* also prompt (some) LLMs to make decisions based on expert decision-making frameworks to self improve.

1

0

5

Rose

@rose_e_wang

6 months

@tiwa_eisape @tallinzen @vansteenkiste_s @feishaAI @mhtessler @MIT @GoogleAI @GoogleDeepMind Super cool work on logical reasoning & biases like ordering effects! One thing I've been thinking about for a while is its connection to the task design lit in math ed (eg ).

2

0

4

Rose

@rose_e_wang

7 days

You want to learn more? Check out Bridge 🌁 in… Paper 📎: Code 💻: Or reach out! I’m happy to chat or talk 🙏

GitHub - rosewang2008/bridge: NAACL 2024. Code & Dataset for "🌁 Bridging the Novice-Expert Gap via...

NAACL 2024. Code & Dataset for "🌁 Bridging the Novice-Expert Gap via Models of Decision-Making: A Case Study on Remediating Math Mistakes" - rosewang2008/bridge

github.com

2

0

5

Rose

@rose_e_wang

7 days

What do we find? Yes, we *can* blend human expert decisions with LLMs and improve. (But still humans > LLMs with expert decision-making > LLMs)

1

0

5

Rose

@rose_e_wang

2 years

We cast this problem as a teacher-student setup where the teacher must first interact to diagnose 🧪the student (the model), before teaching 👩‍🏫(constructing the training dataset). [2/N]

1

0

3

Rose

@rose_e_wang

1 year

@phillip_isola This is super cool!!! I've been exploring a similar idea for math education purposes and there are a lot of cool analogies we use 😀. It's interesting (and ominous??) that at age 30, we seem to hit a dark period though with our analogies

0

3

Rose

@rose_e_wang

3 years

@chelseabfinn @loseydp @DorsaSadigh @corl_conf Congrats everyone!!! 🥳 This is really amazing work!!! 😄

0

3

Rose

@rose_e_wang

7 months

Scaling high-quality tutoring is hard. With growing demand, many platforms use novice tutors who struggle to address mistakes and fail to seize learning opportunities. 📚But, turning mistakes into opportunities is key! Effective strategies can transform student understanding🚀

1

0

3

Rose

@rose_e_wang

7 days

And finally, let’s not forget the amazing friends who gave feedback along the way :) ❤️ @cjziems @JoyHeYueya @GabrielPoesia @chengmyra1 @krisgligoric @malikrali @996roma @lateinteraction A lot of exciting, follow-up work cooking, so … stay tuned 👩‍🍳

1

0

6

Rose

@rose_e_wang

2 years

@ben_eysenbach Hey Ben! Neat work. Using contrastive learning for estimating value functions seems quite similar to the . They study it in symbolic domains + assume env simulator for negatives, so it's slightly different from your domains, but methodologically similar.🙂

1

0

3

Rose

@rose_e_wang

2 months

@AlexTamkin @MicahCarroll Also sharing this on tips for new researchers! Wrote this right after undergrad, so might need to update it with PhD/mentorship reflections 😁

0

3

Rose

@rose_e_wang

2 years

Note: The teacher can’t exhaustively probe the student esp. if (a) the state space is huge [this would take too long!], and (b) with limited communication [eg. if the student were a human, they would get exhausted!] [5/N]

1

0

3

Rose

@rose_e_wang

4 months

Let’s see Edu-ConvoKit in action! 👇 What are examples of real student reasoning in the classroom? Or, examples of how a tutor interacts with a student’s contributions? Boom, with a single function call with Edu-ConvoKit, we view these examples.

1

3

Rose

@rose_e_wang

3 years

Please consider submitting your work to our #ICLR2021 workshop! 😀

Never Ending RL

@neverendingrl

3 years

*** Deadline Extension Alert *** We have decided to extend our @iclr_conf workshop submission deadline by 48 hours, until February 28th AOE. Submit your work here: More details can be found on our website: #ICRL2021 #NERL2021

0

12

26

0

1

3

Rose

@rose_e_wang

4 years

Woops! I got nervous about my first tweet and tagged the wrong Max. Thanks @maxhkw for being an amazing research mentor! :)

0

3

Rose

@rose_e_wang

1 year

Big +1!! You learn how to do sanity checks, debug, and find edge cases through research.. Last summer, I heard an undergrad through my office wall bang on the table, then shout: "Ohhh myyy godddd!!!" *silence* "I'M CHANGING THE LIST AS I'M ITERATING THROUGH IT"😆 what a throwback

Alex Tamkin 🦣

@AlexTamkin

1 year

2) Learn to problem solve Google your errors / stack traces Play around with a library in an interactive python session (Maybe even ask a language model, as long as you can verify the answer it gives you!)

1

0

7

0

2

Rose

@rose_e_wang

7 months

Here’s my favorite qualitative example on a word problem: GPT4 gives away the solution, whereas the teacher suggests an illustrative strategy to help the student understand. Providing the student error and strategy label gets GPT4 to generate pedagogically better responses. 📈

1

0

3

Rose

@rose_e_wang

3 years

And check out our multi-agent cooking environment here! 🍅🥬🍽️

Rose

@rose_e_wang

4 years

Excited to share gym-cooking, a *novel multi-agent Gym environment*: Based on recent work (, #CogSci2020 computational modeling prize winner) with amazing collaborators Sarah Wu, James Evans, Josh Tenenbaum, David Parkes, @maxkw !

6

60

305

0

3

Rose

@rose_e_wang

7 months

@ddmeyer @ddemszky On the other hand, I think this raises Qs abt how we should design collaborative systems/algs when users have very diff preferences in how they want to teach; this diversity Q comes up in other settings too e.g., mental health domain). Would love to chat & ideate about this!

1

0

1

Rose

@rose_e_wang

7 days

@cjziems @JoyHeYueya @GabrielPoesia @chengmyra1 @krisgligoric @malikrali @996roma @lateinteraction I also want to thank @ddmeyer for his insight & inspiring writing e.g., Teachers do so much behind the scenes. That is non-trivial to "chat-bot-away". But, figuring out where NLP can aid educators--that's the exciting challenge I want to focus on!

You Need to Understand That Teachers Process Petabytes Every Day

What would a genAI chatbot do with the class I observed last week?

danmeyer.substack.com

0

4

Rose

@rose_e_wang

7 months

Our work sheds light on the potential and current limitations of using LLMs to provide high-quality learning experiences for tutors & students at scale. We hope that it can serve as a valuable resource for providing equitable, high-quality learning experiences using *real* data.

1

0

3

Rose

@rose_e_wang

2 months

@ecekt2 @eaclmeeting @sandropezzelle @raquel_dmg This is really cool work!! Thanks for an awesome poster session 😃

1

0

3

Rose

@rose_e_wang

7 months

🔥Findings: Expert math teacher>LLMs+expert math teacher’s guidance>LLMs>novice tutors🔥 -The quality of the best LLM’s responses falls short of math teachers. -Providing LLMs w/ teacher’s strategy leads to a 75% improvement in response quality over models w/o that information.

1

0

3