- 17
- 21 756
Generative AI at MIT
เข้าร่วมเมื่อ 16 เม.ย. 2023
Policies & Practice in using AI Writing Assistants
Policies & Practice in using AI Writing Assistants
มุมมอง: 135
วีดีโอ
Human Subject Research: Randomized Controlled Trials
มุมมอง 104ปีที่แล้ว
Human Subject Research: Randomized Controlled Trials
Human Subject Research for LMs: Motivation, Examples, Ethics, Surveys
มุมมอง 230ปีที่แล้ว
Our goal in this module is to summarize practical advice for measuring the impact of your work on people. We’ll cover many subjects, each of which can stand on its own as a whole course! There are links to more reading throughout the slides (available at ai4comm.media.mit.edu/slides/3_8.pdf).
Human Subject Research for LMs: Interviews and Mixed Methods Research
มุมมอง 55ปีที่แล้ว
Human Subject Research for LMs: Interviews and Mixed Methods Research
Programming workshop: Using Python and the OpenAI API to perform and evaluate LLM tasks
มุมมอง 241ปีที่แล้ว
Note: Please see the Github repository for this session at github.com/mit-ccc/MAS-S68-workshop This is a re-recording of our programming workshop, targeted at programmers who want to build and evaluate an LLM-based application with Python and the OpenAI API. In this session we'll focus on writing an application that lets you describe an English word and then guesses which word you mean.. (Some ...
LLM Evaluation Basics: Datasets & Metrics
มุมมอง 10Kปีที่แล้ว
This is an introduction to evaluating Large Language Models (LLMs), which covers what a dataset is, how we measure performance, and how automatic and human evaluation are done.
Evaluation: Bias & Toxicity
มุมมอง 1.2Kปีที่แล้ว
A cursory introduction to the types of work on evaluating bias and fairness in language models and the systems in which they are deployed. (Slides: ai4comm.media.mit.edu/slides/3_1.pdf, starting at slide 22)
Evaluation: Factuality and Halllucination
มุมมอง 746ปีที่แล้ว
In this module we'll look at a few ways that NLP researchers evaluate large language models for factuality (and, relatedly, its evil cousin "hallucination"). (Slides: ai4comm.media.mit.edu/slides/3_1.pdf, starting at slide 31)
Evaluation: LLM robustness and self-consistency
มุมมอง 1.1Kปีที่แล้ว
This video introduces the concepts of robustness and self-consistency in LLMs and describes how to measure them in practice.
John Horton - 4/5 at MIT: Large Language Models as Simulated Economic Agents
มุมมอง 356ปีที่แล้ว
MIT Economist John Horton joined our class to discuss his recent paper Large Language Models as Simulated Economic Agents: What Can We Learn from Homo Silicus?
Rishi Bommasani -- Holistic Evaluation of Language Models (February 15th 2023)
มุมมอง 1.7Kปีที่แล้ว
Rishi Bommasani (PhD student in Computer Science at Stanford) joined MAS.S68 (Generative AI for Constructive Communication) on 2/15 for a lecture on his paper, Holistic Evaluation of Language Models (HELM). HELM paper: arxiv.org/abs/2211.09110 Lecture slides: ai4comm.media.mit.edu/slides/helm.pdf Course website: ai4comm.media.mit.edu/
Prompting large language models - examples
มุมมอง 198ปีที่แล้ว
Lecture discussing some prompting examples of large language models. Prompting examples include grammar error correction, summarization, style transfer, math reasoning, and joke explanation. (Hang Jiang) Slides: ai4comm.media.mit.edu/slides/2_8.pdf
Introduction to prompting large language models
มุมมอง 407ปีที่แล้ว
An introduction to prompting large language models, including prompt tuning. (Shayne Longpre) Slides: ai4comm.media.mit.edu/slides/2_8.pdf
LLM fundamentals - transformers and beyond
มุมมอง 1.2Kปีที่แล้ว
An introduction to the transformers era in NLP and the pretrain/finetune paradigm. (Jad Kabbara) Slides: ai4comm.media.mit.edu/slides/2_8.pdf
Introduction to language models
มุมมอง 461ปีที่แล้ว
In this module we introduce you to language models and show you how they've been applied to products and services we use everyday. We also take a look at the history of language modeling research over the past 30 years, starting from the era of statistical language models and taking you up to the era of large-scale generative transformer-based models. (Doug Beeferman) Slides: ai4comm.media.mit....
Course Intro - Generative AI for Constructive Communication
มุมมอง 570ปีที่แล้ว
Course Intro - Generative AI for Constructive Communication
Jason Wei -- Emergence in Large Language Models (March 1st 2023)
มุมมอง 2.9Kปีที่แล้ว
Jason Wei Emergence in Large Language Models (March 1st 2023)
570 Clotilde Burg
83526 Calista Village
Braun Stravenue
08753 Klein Flat
28977 Lakin Dam
I agree with the other commenter. Also, the flashing toolbar up top was very distracting
Would appreciate if the link to the github repo can be shared
How can I use this package?
IF there is a code demo, it would have helped
wow, thank you for sharing the lectures, specially these that have such a high level of detail and understanding. Me not being in a specific class on this topics, I struggle to find good resources like this. Thank you
give a simple but clear way to explain LLM. Excellent!