The AI Alignment Forum

FEBRUARY 28, 2025

During that time, most of my main research has been motivated by the goal of developing a theoretical foundation for the field of reward learning. The purpose of this sequence is to explain and motivate this research agenda, and to provide an accessible overview over some of my results. And so on.

Research 40

Research Learning Method Policy 40

Beth's Blog: How Nonprofits Can Use Social Media

SEPTEMBER 27, 2006

The learning is not in that moment in which these actions are taking. That's simply because learning is a process that happens when the information shifts from short-term to long-term memory and results in changed beliefs and behaviors. Loretta, doe it depend on which learning theory you buy into?

Knowledge 50

Knowledge Collaboration Social Media Tools 50

Beth's Blog: How Nonprofits Can Use Social Media

JANUARY 3, 2008

The title of this post is a play on a book I read The Book of Learning and Forgetting by Frank Smith in 1998 when I was working with arts educators on integrating technology into their lesson plans. I would recommend technology resources and they would share books about learning. Via email can use up to 5 search terms.

Twitter 50

Twitter Bookmarking RSS Tag 50

The AI Alignment Forum

FEBRUARY 20, 2025

Published on February 20, 2025 11:54 PM GMT TLDR: We made substantial progress in 2024: We published a series of papers that verify key predictions of Singular Learning Theory (SLT) [ 1 , 2 , 3 , 4 , 5 , 6 ]. We scaled key SLT-derived techniques to models with billions of parameters, eliminating our main concerns around tractability.

Technique 52

Technique Sample Structure Training 52

The AI Alignment Forum

FEBRUARY 28, 2025

The results in this paper also make me more interested in multi-objective RL in particular, since it seems quite rich and flexible in terms of what it may enable us to do. In particular, it is not a given that a reasonable incentive structure necessarily can be captured by a reward function.

Learning 40

Learning Discussion Classes Policy 40

Museum 2.0

SEPTEMBER 26, 2012

The purpose of my thesis was two-fold: To research and analyze community and civic engagement practices, methods, theories and examples in other museum programs. To apply the results of my analysis to produce a community-driven program design specifically for implementation at the Santa Cruz Museum of Art & History (the MAH).

Museum 49

Museum Program Community Participatory 49

NonProfit Hub

OCTOBER 16, 2015

Beth is an expert in facilitating online and offline peer learning, curriculum development based on traditional adult learning theory and other instructional approaches. Optimize your search results (they do their Internet research). She has trained thousands of nonprofits around the world. Gen Z by the Numbers.

Social 28

Social Social Media Student Cancer 28

Museum 2.0

JUNE 2, 2010

to find ways that professionals can develop responsive feedback loops where there's an actual demand driving the visitor participation--and a result that changes based on what they do. Basically, the idea is that most organizations learn in a single loop that connects programs to results.

Slides 22

Slides Museum Participatory San Jose 22

The AI Alignment Forum

FEBRUARY 7, 2025

The blue teams goal is to have the resulting system complete the tasks correctly, without being sabotaged. The conditions and results of their experiments resembled the dangerous kind of alignment faking in some ways (e.g., These failure modes are mostly theoretical, but results like Farquhar et al. Greenblatt et al.

Research 52

Research Fund Open Technique 52

DataRobot

APRIL 21, 2022

They define it as “buying” stronger results by just throwing more compute at the model. In the graph below, borrowed from the same article, you can see how some of the most cutting-edge algorithms in deep learning have increased in terms of model size over time. Here, model size is measured by the amount of floating-point operations.

Green 145

Green Environment Metrics Measure 145

Google Research AI blog

NOVEMBER 28, 2022

A Workshop for Algorithmic Efficiency in Practical Neural Network Training Workshop Organizers include: Zachary Nado , George Dahl , Naman Agarwal , Aakanksha Chowdhery Invited Speakers include: Aakanksha Chowdhery , Priya Goyal Human in the Loop Learning (HiLL) Workshop Organizers include: Fisher Yu, Vittorio Ferrari Invited Speakers include: Dorsa (..)

Google 52

Google Language Tutorial Offline 52

The AI Alignment Forum

MARCH 28, 2025

How helpful is it to get these results? Daniel Filan (00:28:50): If people remember my singular learning theory episodes , theyll get mad at you for saying that quadratics are all there is, but its a decent approximation. (00:28:56): Like, the first paper that OpenAI published on GPT interprets GPT.

52

52