The Nonlinear Library: Alignment Forum
By The Nonlinear Fund
Listen to a podcast, please open Podcast Republic app. Available on Google Play Store and Apple App Store.
Image by The Nonlinear Fund
Category: Education
Open in Apple Podcasts
Open RSS feed
Open Website
Rate for this podcast
Subscribers: 2
Reviews: 0
Episodes: 377
Description
The Nonlinear Library allows you to easily listen to top EA and rationalist content on your podcast player. We use text-to-speech software to create an automatically updating repository of audio content from the EA Forum, Alignment Forum, LessWrong, and other EA blogs. To find out more, please visit us at nonlinear.org
| Episode | Date |
|---|---|
|
AF - Simple probes can catch sleeper agents by Monte MacDiarmid
|
Apr 23, 2024 |
|
AF - Dequantifying first-order theories by Jessica Taylor
|
Apr 23, 2024 |
|
AF - ProLU: A Pareto Improvement for Sparse Autoencoders by Glen M. Taggart
|
Apr 23, 2024 |
|
AF - Time complexity for deterministic string machines by alcatal
|
Apr 22, 2024 |
|
AF - Inducing Unprompted Misalignment in LLMs by Sam Svenningsen
|
Apr 19, 2024 |
|
AF - Progress Update #1 from the GDM Mech Interp Team: Full Update by Neel Nanda
|
Apr 19, 2024 |
|
AF - Progress Update #1 from the GDM Mech Interp Team: Summary by Neel Nanda
|
Apr 19, 2024 |
|
AF - Discriminating Behaviorally Identical Classifiers: a model problem for applying interpretability to scalable oversight by Sam Marks
|
Apr 18, 2024 |
|
AF - LLM Evaluators Recognize and Favor Their Own Generations by Arjun Panickssery
|
Apr 17, 2024 |
|
AF - Transformers Represent Belief State Geometry in their Residual Stream by Adam Shai
|
Apr 16, 2024 |
|
AF - Speedrun ruiner research idea by Luke H Miles
|
Apr 13, 2024 |
|
AF - The theory of Proximal Policy Optimisation implementations by salman.mohammadi
|
Apr 12, 2024 |
|
AF - How I select alignment research projects by Ethan Perez
|
Apr 10, 2024 |
|
AF - PIBBSS is hiring in a variety of roles (alignment research and incubation program) by Nora Ammann
|
Apr 09, 2024 |
|
AF - How We Picture Bayesian Agents by johnswentworth
|
Apr 08, 2024 |
|
AF - Measuring Learned Optimization in Small Transformer Models by Jonathan Bostock
|
Apr 08, 2024 |
|
AF - Measuring Predictability of Persona Evaluations by Thee Ho
|
Apr 06, 2024 |
|
AF - Koan: divining alien datastructures from RAM activations by Tsvi Benson-Tilsen
|
Apr 05, 2024 |
|
AF - LLMs for Alignment Research: a safety priority? by Abram Demski
|
Apr 04, 2024 |
|
AF - Run evals on base models too! by orthonormal
|
Apr 04, 2024 |
|
AF - The Case for Predictive Models by Rubi Hudson
|
Apr 03, 2024 |
|
AF - Sparsify: A mechanistic interpretability research agenda by Lee Sharkey
|
Apr 03, 2024 |
|
AF - A Selection of Randomly Selected SAE Features by CallumMcDougall
|
Apr 01, 2024 |
|
AF - SAE-VIS: Announcement Post by CallumMcDougall
|
Mar 31, 2024 |
|
AF - Your LLM Judge may be biased by Rachel Freedman
|
Mar 29, 2024 |
|
AF - SAE reconstruction errors are (empirically) pathological by Wes Gurnee
|
Mar 29, 2024 |
|
AF - How do LLMs give truthful answers? A discussion of LLM vs. human reasoning, ensembles and parrots by Owain Evans
|
Mar 28, 2024 |
|
AF - UDT1.01: The Story So Far (1/10) by Diffractor
|
Mar 27, 2024 |
|
AF - Modern Transformers are AGI, and Human-Level by Abram Demski
|
Mar 26, 2024 |
|
AF - Third-party testing as a key ingredient of AI policy by Zac Hatfield-Dodds
|
Mar 25, 2024 |
|
AF - Announcing Neuronpedia: Platform for accelerating research into Sparse Autoencoders by Johnny Lin
|
Mar 25, 2024 |
|
AF - On the Confusion between Inner and Outer Misalignment by Chris Leong
|
Mar 25, 2024 |
|
AF - Dangers of Closed-Loop AI by Gordon Seidoh Worley
|
Mar 22, 2024 |
|
AF - Video and transcript of presentation on Scheming AIs by Joe Carlsmith
|
Mar 22, 2024 |
|
AF - Comparing Alignment to other AGI interventions: Extensions and analysis by Martín Soto
|
Mar 21, 2024 |
|
AF - Stagewise Development in Neural Networks by Jesse Hoogland
|
Mar 20, 2024 |
|
AF - Natural Latents: The Concepts by johnswentworth
|
Mar 20, 2024 |
|
AF - Comparing Alignment to other AGI interventions: Basic model by Martín Soto
|
Mar 20, 2024 |
|
AF - New report: Safety Cases for AI by Josh Clymer
|
Mar 20, 2024 |
|
AF - AtP*: An efficient and scalable method for localizing LLM behaviour to components by Neel Nanda
|
Mar 18, 2024 |
|
AF - Improving SAE's by Sqrt()-ing L1 and Removing Lowest Activating Features by Logan Riggs Smith
|
Mar 15, 2024 |
|
AF - More people getting into AI safety should do a PhD by AdamGleave
|
Mar 14, 2024 |
|
AF - Laying the Foundations for Vision and Multimodal Mechanistic Interpretability and Open Problems by Sonia Joseph
|
Mar 13, 2024 |
|
AF - Virtual AI Safety Unconference 2024 by Orpheus Lummis
|
Mar 13, 2024 |
|
AF - Transformer Debugger by Henk Tillman
|
Mar 12, 2024 |
|
AF - Open consultancy: Letting untrusted AIs choose what answer to argue for by Fabien Roger
|
Mar 12, 2024 |
|
AF - Bias-Augmented Consistency Training Reduces Biased Reasoning in Chain-of-Thought by miles
|
Mar 11, 2024 |
|
AF - How disagreements about Evidential Correlations could be settled by Martín Soto
|
Mar 11, 2024 |
|
AF - Understanding SAE Features with the Logit Lens by Joseph Isaac Bloom
|
Mar 11, 2024 |
|
AF - 0th Person and 1st Person Logic by Adele Lopez
|
Mar 10, 2024 |
|
AF - Scenario Forecasting Workshop: Materials and Learnings by elifland
|
Mar 08, 2024 |
|
AF - Forecasting future gains due to post-training enhancements by elifland
|
Mar 08, 2024 |
|
AF - Evidential Correlations are Subjective, and it might be a problem by Martín Soto
|
Mar 07, 2024 |
|
AF - We Inspected Every Head In GPT-2 Small using SAEs So You Don't Have To by robertzk
|
Mar 06, 2024 |
|
AF - Many arguments for AI x-risk are wrong by Alex Turner
|
Mar 05, 2024 |
|
AF - Anthropic release Claude 3, claims >GPT-4 Performance by Lawrence Chan
|
Mar 04, 2024 |
|
AF - Some costs of superposition by Linda Linsefors
|
Mar 03, 2024 |
|
AF - Approaching Human-Level Forecasting with Language Models by Fred Zhang
|
Feb 29, 2024 |
|
AF - Tips for Empirical Alignment Research by Ethan Perez
|
Feb 29, 2024 |
|
AF - Post series on "Liability Law for reducing Existential Risk from AI" by Nora Ammann
|
Feb 29, 2024 |
|
AF - Timaeus's First Four Months by Jesse Hoogland
|
Feb 28, 2024 |
|
AF - Notes on control evaluations for safety cases by Ryan Greenblatt
|
Feb 28, 2024 |
|
AF - Counting arguments provide no evidence for AI doom by Nora Belrose
|
Feb 27, 2024 |
|
AF - Deconfusing In-Context Learning by Arjun Panickssery
|
Feb 25, 2024 |
|
AF - Instrumental deception and manipulation in LLMs - a case study by Olli Järviniemi
|
Feb 24, 2024 |
|
AF - The Shutdown Problem: Incomplete Preferences as a Solution by Elliott Thornley
|
Feb 23, 2024 |
|
AF - Analogies between scaling labs and misaligned superintelligent AI by Stephen Casper
|
Feb 21, 2024 |
|
AF - Extinction Risks from AI: Invisible to Science? by Vojtech Kovarik
|
Feb 21, 2024 |
|
AF - Extinction-level Goodhart's Law as a Property of the Environment by Vojtech Kovarik
|
Feb 21, 2024 |
|
AF - Dynamics Crucial to AI Risk Seem to Make for Complicated Models by Vojtech Kovarik
|
Feb 21, 2024 |
|
AF - Which Model Properties are Necessary for Evaluating an Argument? by Vojtech Kovarik
|
Feb 21, 2024 |
|
AF - Weak vs Quantitative Extinction-level Goodhart's Law by Vojtech Kovarik
|
Feb 21, 2024 |
|
AF - Why does generalization work? by Martín Soto
|
Feb 20, 2024 |
|
AF - Complexity classes for alignment properties by Arun Jose
|
Feb 20, 2024 |
|
AF - Protocol evaluations: good analogies vs control by Fabien Roger
|
Feb 19, 2024 |
|
AF - Self-Awareness: Taxonomy and eval suite proposal by Daniel Kokotajlo
|
Feb 17, 2024 |
|
AF - The Pointer Resolution Problem by Arun Jose
|
Feb 16, 2024 |
|
AF - Retrospective: PIBBSS Fellowship 2023 by DusanDNesic
|
Feb 16, 2024 |
|
AF - Searching for Searching for Search by Rubi Hudson
|
Feb 14, 2024 |
|
AF - Critiques of the AI control agenda by Arun Jose
|
Feb 14, 2024 |
|
AF - Requirements for a Basin of Attraction to Alignment by Roger Dearnaley
|
Feb 14, 2024 |
|
AF - Interpreting Quantum Mechanics in Infra-Bayesian Physicalism by Yegreg
|
Feb 12, 2024 |
|
AF - Natural abstractions are observer-dependent: a conversation with John Wentworth by Martín Soto
|
Feb 12, 2024 |
|
AF - Updatelessness doesn't solve most problems by Martín Soto
|
Feb 08, 2024 |
|
AF - Debating with More Persuasive LLMs Leads to More Truthful Answers by Akbir Khan
|
Feb 07, 2024 |
|
AF - How to train your own "Sleeper Agents" by Evan Hubinger
|
Feb 07, 2024 |
|
AF - what does davidad want from "boundaries"? by Chipmonk
|
Feb 06, 2024 |
|
AF - Preventing exfiltration via upload limits seems promising by Ryan Greenblatt
|
Feb 06, 2024 |
|
AF - Attention SAEs Scale to GPT-2 Small by Connor Kissane
|
Feb 03, 2024 |
|
AF - Survey for alignment researchers: help us build better field-level models by Cameron Berg
|
Feb 02, 2024 |
|
AF - Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small by Joseph Isaac Bloom
|
Feb 02, 2024 |
|
AF - Evaluating Stability of Unreflective Alignment by james.lucassen
|
Feb 01, 2024 |
|
AF - PIBBSS Speaker events comings up in February by DusanDNesic
|
Feb 01, 2024 |
|
AF - Last call for submissions for TAIS 2024! by Blaine William Rogers
|
Jan 30, 2024 |
|
AF - The case for more ambitious language model evals by Arun Jose
|
Jan 30, 2024 |
|
AF - Agents that act for reasons: a thought experiment by Michele Campolo
|
Jan 24, 2024 |
|
AF - We need a science of evals by Marius Hobbhahn
|
Jan 22, 2024 |
|
AF - InterLab - a toolkit for experiments with multi-agent interactions by Tomáš Gavenčiak
|
Jan 22, 2024 |
|
AF - A Shutdown Problem Proposal by johnswentworth
|
Jan 21, 2024 |
|
AF - Four visions of Transformative AI success by Steve Byrnes
|
Jan 17, 2024 |
|
AF - Managing catastrophic misuse without robust AIs by Ryan Greenblatt
|
Jan 16, 2024 |
|
AF - Sparse Autoencoders Work on Attention Layer Outputs by Connor Kissane
|
Jan 16, 2024 |
|
AF - Investigating Bias Representations in LLMs via Activation Steering by DawnLu
|
Jan 15, 2024 |
|
AF - Goals selected from learned knowledge: an alternative to RL alignment by Seth Herd
|
Jan 15, 2024 |
|
AF - Three Types of Constraints in the Space of Agents by Nora Ammann
|
Jan 15, 2024 |
|
AF - Introducing Alignment Stress-Testing at Anthropic by Evan Hubinger
|
Jan 12, 2024 |
|
AF - Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training by Evan Hubinger
|
Jan 12, 2024 |
|
AF - Apply to the PIBBSS Summer Research Fellowship by Nora Ammann
|
Jan 12, 2024 |
|
AF - Goodbye, Shoggoth: The Stage, its Animatronics, and the Puppeteer - a New Metaphor by Roger Dearnaley
|
Jan 09, 2024 |
|
AF - A starter guide for evals by Marius Hobbhahn
|
Jan 08, 2024 |
|
AF - Deceptive AI Deceptively-aligned AI by Steve Byrnes
|
Jan 07, 2024 |
|
AF - Catching AIs red-handed by Ryan Greenblatt
|
Jan 05, 2024 |
|
AF - Predictive model agents are sort of corrigible by Raymond D
|
Jan 05, 2024 |
|
AF - What's up with LLMs representing XORs of arbitrary features? by Sam Marks
|
Jan 03, 2024 |
|
AF - Safety First: safety before full alignment. The deontic sufficiency hypothesis. by Chipmonk
|
Jan 03, 2024 |
|
AF - Steering Llama-2 with contrastive activation additions by Nina Rimsky
|
Jan 02, 2024 |
|
AF - Mech Interp Challenge: January - Deciphering the Caesar Cipher Model by CallumMcDougall
|
Jan 01, 2024 |
|
AF - A hermeneutic net for agency by Tsvi Benson-Tilsen
|
Jan 01, 2024 |
|
AF - A case for AI alignment being difficult by Jessica Taylor
|
Dec 31, 2023 |
|
AF - AI Alignment Metastrategy by Vanessa Kosoy
|
Dec 31, 2023 |
|
AF - Free agents by Michele Campolo
|
Dec 27, 2023 |
|
AF - Critical review of Christiano's disagreements with Yudkowsky by Vanessa Kosoy
|
Dec 27, 2023 |
|
AF - AGI will be made of heterogeneous components, Transformer and Selective SSM blocks will be among them by Roman Leventov
|
Dec 27, 2023 |
|
AF - 5. Moral Value for Sentient Animals? Alas, Not Yet by Roger Dearnaley
|
Dec 27, 2023 |
|
AF - Fact Finding: Do Early Layers Specialise in Local Processing? (Post 5) by Neel Nanda
|
Dec 23, 2023 |
|
AF - Measurement tampering detection as a special case of weak-to-strong generalization by Ryan Greenblatt
|
Dec 23, 2023 |
|
AF - Idealized Agents Are Approximate Causal Mirrors (+ Radical Optimism on Agent Foundations) by Thane Ruthenis
|
Dec 22, 2023 |
|
AF - Open positions: Research Analyst at the AI Standards Lab by Koen Holtman
|
Dec 22, 2023 |
|
AF - How Would an Utopia-Maximizer Look Like? by Thane Ruthenis
|
Dec 20, 2023 |
|
AF - Meaning and Agency by Abram Demski
|
Dec 19, 2023 |
|
AF - Don't Share Information Exfohazardous on Others' AI-Risk Models by Thane Ruthenis
|
Dec 19, 2023 |
|
AF - Paper: Tell, Don't Show- Declarative facts influence how LLMs generalize by Owain Evans
|
Dec 19, 2023 |
|
AF - Assessment of AI safety agendas: think about the downside risk by Roman Leventov
|
Dec 19, 2023 |
|
AF - The Shortest Path Between Scylla and Charybdis by Thane Ruthenis
|
Dec 18, 2023 |
|
AF - Discussion: Challenges with Unsupervised LLM Knowledge Discovery by Seb Farquhar
|
Dec 18, 2023 |
|
AF - Interpreting the Learning of Deceit by Roger Dearnaley
|
Dec 18, 2023 |
|
AF - A Common-Sense Case For Mutually-Misaligned AGIs Allying Against Humans by Thane Ruthenis
|
Dec 17, 2023 |
|
AF - OpenAI, DeepMind, Anthropic, etc. should shut down. by Tamsin Leake
|
Dec 17, 2023 |
|
AF - Bounty: Diverse hard tasks for LLM agents by Beth Barnes
|
Dec 17, 2023 |
|
AF - Scalable Oversight and Weak-to-Strong Generalization: Compatible approaches to the same problem by Ansh Radhakrishnan
|
Dec 16, 2023 |
|
AF - Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision by leogao
|
Dec 16, 2023 |
|
AF - Current AIs Provide Nearly No Data Relevant to AGI Alignment by Thane Ruthenis
|
Dec 15, 2023 |
|
AF - AI Control: Improving Safety Despite Intentional Subversion by Buck Shlegeris
|
Dec 13, 2023 |
|
AF - Some biases and selection effects in AI risk discourse by Tamsin Leake
|
Dec 12, 2023 |
|
AF - Adversarial Robustness Could Help Prevent Catastrophic Misuse by Aidan O'Gara
|
Dec 11, 2023 |
|
AF - Empirical work that might shed light on scheming (Section 6 of "Scheming AIs") by Joe Carlsmith
|
Dec 11, 2023 |
|
AF - Quick thoughts on the implications of multi-agent views of mind on AI takeover by Kaj Sotala
|
Dec 11, 2023 |
|
AF - Auditing failures vs concentrated failures by Ryan Greenblatt
|
Dec 11, 2023 |
|
AF - How LDT helps reduce the AI arms race by Tamsin Leake
|
Dec 10, 2023 |
|
AF - Send us example gnarly bugs by Beth Barnes
|
Dec 10, 2023 |
|
AF - Summing up "Scheming AIs" (Section 5) by Joe Carlsmith
|
Dec 09, 2023 |
|
AF - Finding Sparse Linear Connections between Features in LLMs by Logan Riggs Smith
|
Dec 09, 2023 |
|
AF - Speed arguments against scheming (Section 4.4-4.7 of "Scheming AIs") by Joe Carlsmith
|
Dec 08, 2023 |
|
AF - Simplicity arguments for scheming (Section 4.3 of "Scheming AIs") by Joe Carlsmith
|
Dec 07, 2023 |
|
AF - The counting argument for scheming (Sections 4.1 and 4.2 of "Scheming AIs") by Joe Carlsmith
|
Dec 06, 2023 |
|
AF - Google Gemini Announced by g-w1
|
Dec 06, 2023 |
|
AF - Arguments for/against scheming that focus on the path SGD takes (Section 3 of "Scheming AIs") by Joe Carlsmith
|
Dec 05, 2023 |
|
AF - Studying The Alien Mind by Quentin Feuillade--Montixi
|
Dec 05, 2023 |
|
AF - Deep Forgetting and Unlearning for Safely-Scoped LLMs by Stephen Casper
|
Dec 05, 2023 |
|
AF - Neural uncertainty estimation for alignment by Charlie Steiner
|
Dec 05, 2023 |
|
AF - Some open-source dictionaries and dictionary learning infrastructure by Sam Marks
|
Dec 05, 2023 |
|
AF - 2023 Alignment Research Updates from FAR AI by AdamGleave
|
Dec 04, 2023 |
|
AF - What's new at FAR AI by AdamGleave
|
Dec 04, 2023 |
|
AF - Non-classic stories about scheming (Section 2.3.2 of "Scheming AIs") by Joe Carlsmith
|
Dec 04, 2023 |
|
AF - Does scheming lead to adequate future empowerment? (Section 2.3.1.2 of "Scheming AIs") by Joe Carlsmith
|
Dec 03, 2023 |
|
AF - The goal-guarding hypothesis (Section 2.3.1.1 of "Scheming AIs") by Joe Carlsmith
|
Dec 02, 2023 |
|
AF - Thoughts on "AI is easy to control" by Pope and Belrose by Steve Byrnes
|
Dec 01, 2023 |
|
AF - How useful for alignment-relevant work are AIs with short-term goals? (Section 2.2.4.3 of "Scheming AIs") by Joe Carlsmith
|
Dec 01, 2023 |
|
AF - FixDT by Abram Demski
|
Nov 30, 2023 |
|
AF - Is scheming more likely in models trained to have long-term goals? (Sections 2.2.4.1-2.2.4.2 of "Scheming AIs") by Joe Carlsmith
|
Nov 30, 2023 |
|
AF - [Linkpost] Remarks on the Convergence in Distribution of Random Neural Networks to Gaussian Processes in the Infinite Width Limit by Spencer Becker-Kahn
|
Nov 30, 2023 |
|
AF - "Clean" vs. "messy" goal-directedness (Section 2.2.3 of "Scheming AIs") by Joe Carlsmith
|
Nov 29, 2023 |
|
AF - Intro to Superposition and Sparse Autoencoders (Colab exercises) by CallumMcDougall
|
Nov 29, 2023 |
|
AF - How to Control an LLM's Behavior (why my P(DOOM) went down) by Roger Dearnaley
|
Nov 28, 2023 |
|
AF - Two sources of beyond-episode goals (Section 2.2.2 of "Scheming AIs") by Joe Carlsmith
|
Nov 28, 2023 |
|
AF - Anthropic Fall 2023 Debate Progress Update by Ansh Radhakrishnan
|
Nov 28, 2023 |
|
AF - AISC 2024 - Project Summaries by Nicky Pochinkov
|
Nov 27, 2023 |
|
AF - There is no IQ for AI by Gabriel Alfour
|
Nov 27, 2023 |
|
AF - Two concepts of an "episode" (Section 2.2.1 of "Scheming AIs") by Joe Carlsmith
|
Nov 27, 2023 |
|
AF - Situational awareness (Section 2.1 of "Scheming AIs") by Joe Carlsmith
|
Nov 26, 2023 |
|
AF - On "slack" in training (Section 1.5 of "Scheming AIs") by Joe Carlsmith
|
Nov 25, 2023 |
|
AF - Why focus on schemers in particular (Sections 1.3 and 1.4 of "Scheming AIs") by Joe Carlsmith
|
Nov 24, 2023 |
|
AF - Ability to solve long-horizon tasks correlates with wanting things in the behaviorist sense by Nate Soares
|
Nov 24, 2023 |
|
AF - 4. A Moral Case for Evolved-Sapience-Chauvinism by Roger Dearnaley
|
Nov 24, 2023 |
|
AF - 3. Uploading by Roger Dearnaley
|
Nov 23, 2023 |
|
AF - Thomas Kwa's research journal by Thomas Kwa
|
Nov 23, 2023 |
|
AF - A taxonomy of non-schemer models (Section 1.2 of "Scheming AIs") by Joe Carlsmith
|
Nov 22, 2023 |
|
AF - Public Call for Interest in Mathematical Alignment by David Manheim
|
Nov 22, 2023 |
|
AF - Varieties of fake alignment (Section 1.1 of "Scheming AIs") by Joe Carlsmith
|
Nov 21, 2023 |
|
AF - Alignment can improve generalisation through more robustly doing what a human wants - CoinRun example by Stuart Armstrong
|
Nov 21, 2023 |
|
AF - Agent Boundaries Aren't Markov Blankets. by Abram Demski
|
Nov 20, 2023 |
|
AF - New paper shows truthfulness and instruction-following don't generalize by default by Josh Clymer
|
Nov 19, 2023 |
|
AF - My Criticism of Singular Learning Theory by Joar Skalse
|
Nov 19, 2023 |
|
AF - AI Safety Camp 2024 by Linda Linsefors
|
Nov 18, 2023 |
|
AF - Sam Altman fired from OpenAI by Lawrence Chan
|
Nov 17, 2023 |
|
AF - Coup probes trained off-policy by Fabien Roger
|
Nov 17, 2023 |
|
AF - Evaluating AI Systems for Moral Status Using Self-Reports by Ethan Perez
|
Nov 16, 2023 |
|
AF - Experiences and learnings from both sides of the AI safety job market by Marius Hobbhahn
|
Nov 15, 2023 |
|
AF - Theories of Change for AI Auditing by Lee Sharkey
|
Nov 13, 2023 |
|
AF - Open Phil releases RFPs on LLM Benchmarks and Forecasting by Lawrence Chan
|
Nov 11, 2023 |
|
AF - We have promising alignment plans with low taxes by Seth Herd
|
Nov 10, 2023 |
|
AF - Learning-theoretic agenda reading list by Vanessa Kosoy
|
Nov 09, 2023 |
|
AF - Five projects from AI Safety Hub Labs 2023 by Charlie Griffin
|
Nov 08, 2023 |
|
AF - Tall Tales at Different Scales: Evaluating Scaling Trends For Deception In Language Models by Felix Hofstätter
|
Nov 08, 2023 |
|
AF - Scalable And Transferable Black-Box Jailbreaks For Language Models Via Persona Modulation by Soroush Pour
|
Nov 07, 2023 |
|
AF - Box inversion revisited by Jan Kulveit
|
Nov 07, 2023 |
|
AF - Announcing TAIS 2024 by Blaine William Rogers
|
Nov 06, 2023 |
|
AF - Genetic fitness is a measure of selection strength, not the selection target by Kaj Sotala
|
Nov 04, 2023 |
|
AF - Untrusted smart models and trusted dumb models by Buck Shlegeris
|
Nov 04, 2023 |
|
AF - Thoughts on open source AI by Sam Marks
|
Nov 03, 2023 |
|
AF - Mech Interp Challenge: November - Deciphering the Cumulative Sum Model by TheMcDouglas
|
Nov 02, 2023 |
|
AF - My thoughts on the social response to AI risk by Matthew Barnett
|
Nov 01, 2023 |
|
AF - Dario Amodei's prepared remarks from the UK AI Safety Summit, on Anthropic's Responsible Scaling Policy by Zac Hatfield-Dodds
|
Nov 01, 2023 |
|
AF - 4. Risks from causing illegitimate value change (performative predictors) by Nora Ammann
|
Oct 26, 2023 |
|
AF - 3. Premise three and Conclusion: AI systems can affect value change trajectories and the Value Change Problem by Nora Ammann
|
Oct 26, 2023 |
|
AF - I don't find the lie detection results that surprising (by an author of the paper) by JanBrauner
|
Oct 04, 2023 |
|
AF - Some Quick Follow-Up Experiments to "Taken out of context: On measuring situational awareness in LLMs" by miles
|
Oct 03, 2023 |
|
AF - Direction of Fit by Nicholas Kees Dupuis
|
Oct 02, 2023 |
|
AF - New Tool: the Residual Stream Viewer by Adam Yedidia
|
Oct 01, 2023 |
|
AF - How model editing could help with the alignment problem by Michael Ripa
|
Sep 30, 2023 |
|
AF - How to Catch an AI Liar: Lie Detection in Black-Box LLMs by Asking Unrelated Questions by JanBrauner
|
Sep 28, 2023 |
|
AF - Alignment Workshop talks by Richard Ngo
|
Sep 28, 2023 |
|
AF - Different views of alignment have different consequences for imperfect methods by Stuart Armstrong
|
Sep 28, 2023 |
|
AF - Projects I would like to see (possibly at AI Safety Camp) by Linda Linsefors
|
Sep 27, 2023 |
|
AF - Announcing the CNN Interpretability Competition by Stephen Casper
|
Sep 26, 2023 |
|
AF - Impact stories for model internals: an exercise for interpretability researchers by Jenny Nitishinskaya
|
Sep 25, 2023 |
|
AF - What is wrong with this "utility switch button problem" approach? by Donald Hobson
|
Sep 25, 2023 |
|
AF - Understanding strategic deception and deceptive alignment by Marius Hobbhahn
|
Sep 25, 2023 |
|
AF - Sparse Autoencoders: Future Work by Logan Riggs Smith
|
Sep 21, 2023 |
|
AF - Sparse Autoencoders Find Highly Interpretable Directions in Language Models by Logan Riggs Smith
|
Sep 21, 2023 |
|
AF - There should be more AI safety orgs by Marius Hobbhahn
|
Sep 21, 2023 |
|
AF - Image Hijacks: Adversarial Images can Control Generative Models at Runtime by Scott Emmons
|
Sep 20, 2023 |
|
AF - Interpretability Externalities Case Study - Hungry Hungry Hippos by Magdalena Wache
|
Sep 20, 2023 |
|
AF - Anthropic's Responsible Scaling Policy and Long Term Benefit Trust by Zac Hatfield-Dodds
|
Sep 19, 2023 |
|
AF - Where might I direct promising-to-me researchers to apply for alignment jobs/grants? by Abram Demski
|
Sep 18, 2023 |
|
AF - Three ways interpretability could be impactful by Arthur Conmy
|
Sep 18, 2023 |
|
AF - Telopheme, telophore, and telotect by Tsvi Benson-Tilsen
|
Sep 17, 2023 |
|
AF - How to talk about reasons why AGI might not be near? by Kaj Sotala
|
Sep 17, 2023 |
|
AF - Uncovering Latent Human Wellbeing in LLM Embeddings by ChengCheng
|
Sep 14, 2023 |
|
AF - Mech Interp Challenge: September - Deciphering the Addition Model by TheMcDouglas
|
Sep 13, 2023 |
|
AF - Apply to lead a project during the next virtual AI Safety Camp by Linda Linsefors
|
Sep 13, 2023 |
|
AF - UDT shows that decision theory is more puzzling than ever by Wei Dai
|
Sep 13, 2023 |
|
AF - Focus on the Hardest Part First by Johannes C. Mayer
|
Sep 11, 2023 |
|
AF - Explaining grokking through circuit efficiency by Vikrant Varma
|
Sep 08, 2023 |
|
AF - The Löbian Obstacle, And Why You Should Care by marc/er
|
Sep 07, 2023 |
|
AF - Recreating the caring drive by Catnee
|
Sep 07, 2023 |
|
AF - ActAdd: Steering Language Models without Optimization by technicalities
|
Sep 06, 2023 |
|
AF - What I would do if I wasn't at ARC Evals by Lawrence Chan
|
Sep 05, 2023 |
|
AF - Benchmarks for Detecting Measurement Tampering [Redwood Research] by Ryan Greenblatt
|
Sep 05, 2023 |
|
AF - Paper: On measuring situational awareness in LLMs by Owain Evans
|
Sep 04, 2023 |
|
AF - Fundamental question: What determines a mind's effects? by Tsvi Benson-Tilsen
|
Sep 03, 2023 |
|
AF - Series of absurd upgrades in nature's great search by Luke H Miles
|
Sep 03, 2023 |
|
AF - PIBBSS Summer Symposium 2023 by Nora Ammann
|
Sep 02, 2023 |
|
AF - Tensor Trust: An online game to uncover prompt injection vulnerabilities by Luke Bailey
|
Sep 01, 2023 |
|
AF - Meta Questions about Metaphilosophy by Wei Dai
|
Sep 01, 2023 |
|
AF - Responses to apparent rationalist confusions about game / decision theory by Anthony DiGiovanni
|
Aug 30, 2023 |
|
AF - Paper Walkthrough: Automated Circuit Discovery with Arthur Conmy by Neel Nanda
|
Aug 29, 2023 |
|
AF - An OV-Coherent Toy Model of Attention Head Superposition by LaurenGreenspan
|
Aug 29, 2023 |
|
AF - Barriers to Mechanistic Interpretability for AGI Safety by Connor Leahy
|
Aug 29, 2023 |
|
AF - AI Deception: A Survey of Examples, Risks, and Potential Solutions by Simon Goldstein
|
Aug 29, 2023 |
|
AF - OpenAI base models are not sycophantic, at any size by nostalgebraist
|
Aug 29, 2023 |
|
AF - Paradigms and Theory Choice in AI: Adaptivity, Economy and Control by particlemania
|
Aug 28, 2023 |
|
AF - A list of core AI safety problems and how I hope to solve them by davidad (David A. Dalrymple)
|
Aug 26, 2023 |
|
AF - Red-teaming language models via activation engineering by Nina Rimsky
|
Aug 26, 2023 |
|
AF - A Model-based Approach to AI Existential Risk by Samuel Dylan Martin
|
Aug 25, 2023 |
|
AF - Implications of evidential cooperation in large worlds by Lukas Finnveden
|
Aug 23, 2023 |
|
AF - Causality and a Cost Semantics for Neural Networks by scottviteri
|
Aug 21, 2023 |
|
AF - "Dirty concepts" in AI alignment discourses, and some guesses for how to deal with them by Nora Ammann
|
Aug 20, 2023 |
|
AF - We can do better than DoWhatIMean by Luke H Miles
|
Aug 19, 2023 |
|
AF - An Overview of Catastrophic AI Risks: Summary by Dan H
|
Aug 18, 2023 |
|
AF - Managing risks of our own work by Beth Barnes
|
Aug 18, 2023 |
|
AF - Autonomous replication and adaptation: an attempt at a concrete danger threshold by Hjalmar Wijk
|
Aug 17, 2023 |
|
AF - If we had known the atmosphere would ignite by Jeffs
|
Aug 16, 2023 |
|
AF - A Proof of Löb's Theorem using Computability Theory by Jessica Taylor
|
Aug 16, 2023 |
|
AF - AGI is easier than robotaxis by Daniel Kokotajlo
|
Aug 13, 2023 |
|
AF - When discussing AI risks, talk about capabilities, not intelligence by Victoria Krakovna
|
Aug 11, 2023 |
|
AF - Linkpost: We need another Expert Survey on Progress in AI, urgently by David Mears
|
Aug 11, 2023 |
|
AF - Could We Automate AI Alignment Research? by Stephen McAleese
|
Aug 10, 2023 |
|
AF - The positional embedding matrix and previous-token heads: how do they actually work? by Adam Yedidia
|
Aug 10, 2023 |
|
AF - Mech Interp Challenge: August - Deciphering the First Unique Character Model by TheMcDouglas
|
Aug 09, 2023 |
|
AF - Ground-Truth Label Imbalance Impairs Contrast-Consistent Search Performance by Tom Angsten
|
Aug 09, 2023 |
|
AF - Modulating sycophancy in an RLHF model via activation steering by NinaR
|
Aug 09, 2023 |
|
AF - Model Organisms of Misalignment: The Case for a New Pillar of Alignment Research by Evan Hubinger
|
Aug 08, 2023 |
|
AF - An interactive introduction to grokking and mechanistic interpretability by Adam Pearce
|
Aug 07, 2023 |
|
AF - Yann LeCun on AGI and AI Safety by Chris Leong
|
Aug 06, 2023 |
|
AF - Password-locked models: a stress case for capabilities evaluation by Fabien Roger
|
Aug 03, 2023 |
|
AF - 3 levels of threat obfuscation by HoldenKarnofsky
|
Aug 02, 2023 |
|
AF - ARC Evals new report: Evaluating Language-Model Agents on Realistic Autonomous Tasks by Beth Barnes
|
Aug 01, 2023 |
|
AF - The "no sandbagging on checkable tasks" hypothesis by Joe Carlsmith
|
Jul 31, 2023 |
|
AF - Watermarking considered overrated? by DanielFilan
|
Jul 31, 2023 |
|
AF - Thoughts on sharing information about language model capabilities by Paul Christiano
|
Jul 31, 2023 |
|
AF - Open Problems and Fundamental Limitations of RLHF by Stephen Casper
|
Jul 31, 2023 |
|
AF - When can we trust model evaluations? by Evan Hubinger
|
Jul 28, 2023 |
|
AF - Reducing sycophancy and improving honesty via activation steering by NinaR
|
Jul 28, 2023 |
|
AF - Mech Interp Puzzle 2: Word2Vec Style Embeddings by Neel Nanda
|
Jul 28, 2023 |
|
AF - Meta-level adversarial evaluation of oversight techniques might allow robust measurement of their adequacy by Buck Shlegeris
|
Jul 26, 2023 |
|
AF - Frontier Model Security by Matthew "Vaniver" Gray
|
Jul 26, 2023 |
|
AF - How LLMs are and are not myopic by janus
|
Jul 25, 2023 |
|
AF - Open problems in activation engineering by Alex Turner
|
Jul 24, 2023 |
|
AF - QAPR 5: grokking is maybe not that big a deal? by Quintin Pope
|
Jul 23, 2023 |
|
AF - Examples of Prompts that Make GPT-4 Output Falsehoods by Stephen Casper
|
Jul 22, 2023 |
|
AF - Reward Hacking from a Causal Perspective by Tom Everitt
|
Jul 21, 2023 |
|
AF - Priorities for the UK Foundation Models Taskforce by Andrea Miotti
|
Jul 21, 2023 |
|
AF - Even Superhuman Go AIs Have Surprising Failures Modes by AdamGleave
|
Jul 20, 2023 |
|
AF - Does Circuit Analysis Interpretability Scale? Evidence from Multiple Choice Capabilities in Chinchilla by Neel Nanda
|
Jul 20, 2023 |
|
AF - Speculative inferences about path dependence in LLM supervised fine-tuning from results on linear mode connectivity and model souping by Robert Kirk
|
Jul 20, 2023 |
|
AF - Alignment Grantmaking is Funding-Limited Right Now by johnswentworth
|
Jul 19, 2023 |
|
AF - Tiny Mech Interp Projects: Emergent Positional Embeddings of Words by Neel Nanda
|
Jul 18, 2023 |
|
AF - Still no Lie Detector for LLMs by Daniel Herrmann
|
Jul 18, 2023 |
|
AF - Meta announces Llama 2; "open sources" it for commercial use by Lawrence Chan
|
Jul 18, 2023 |
|
AF - Measuring and Improving the Faithfulness of Model-Generated Reasoning by Ansh Radhakrishnan
|
Jul 18, 2023 |
|
AF - Thoughts on "Process-Based Supervision" by Steve Byrnes
|
Jul 17, 2023 |
|
AF - AutoInterpretation Finds Sparse Coding Beats Alternatives by Hoagy
|
Jul 17, 2023 |
|
AF - Mech Interp Puzzle 1: Suspiciously Similar Embeddings in GPT-Neo by Neel Nanda
|
Jul 16, 2023 |
|
AF - Robustness of Model-Graded Evaluations and Automated Interpretability by Simon Lermen
|
Jul 15, 2023 |
|
AF - Eric Michaud on the Quantization Model of Neural Scaling, Interpretability and Grokking by Michaël Trazzi
|
Jul 12, 2023 |
|
AF - What does the launch of x.ai mean for AI Safety? by Chris Leong
|
Jul 12, 2023 |
|
AF - Towards Developmental Interpretability by Jesse Hoogland
|
Jul 12, 2023 |
|
AF - Goal-Direction for Simulated Agents by Raymond D
|
Jul 12, 2023 |
|
AF - Incentives from a causal perspective by Tom Everitt
|
Jul 10, 2023 |
|
AF - “Reframing Superintelligence” + LLMs + 4 years by Eric Drexler
|
Jul 10, 2023 |
|
AF - Open-minded updatelessness by Nicolas Macé
|
Jul 10, 2023 |
|
AF - Consciousness as a conflationary alliance term by Andrew Critch
|
Jul 10, 2023 |
|
AF - Really Strong Features Found in Residual Stream by Logan Riggs Smith
|
Jul 08, 2023 |
|
AF - Seven Strategies for Tackling the Hard Part of the Alignment Problem by Stephen Casper
|
Jul 08, 2023 |
|
AF - Continuous Adversarial Quality Assurance: Extending RLHF and Constitutional AI by Benaya Koren
|
Jul 08, 2023 |
|
AF - "Concepts of Agency in Biology" (Okasha, 2023) - Brief Paper Summary by Nora Ammann
|
Jul 08, 2023 |
|
AF - Views on when AGI comes and on strategy to reduce existential risk by Tsvi Benson-Tilsen
|
Jul 08, 2023 |
|
AF - Jesse Hoogland on Developmental Interpretability and Singular Learning Theory by Michaël Trazzi
|
Jul 06, 2023 |
|
AF - [Linkpost] Introducing Superalignment by Beren Millidge
|
Jul 05, 2023 |
|
AF - (tentatively) Found 600+ Monosemantic Features in a Small LM Using Sparse Autoencoders by Logan Riggs Smith
|
Jul 05, 2023 |
|
AF - Ten Levels of AI Alignment Difficulty by Samuel Dylan Martin
|
Jul 03, 2023 |
|
AF - VC Theory Overview by Joar Skalse
|
Jul 02, 2023 |
|
AF - Sources of evidence in Alignment by Martín Soto
|
Jul 02, 2023 |
|
AF - Quantitative cruxes in Alignment by Martín Soto
|
Jul 02, 2023 |
|
AF - How Smart Are Humans? by Joar Skalse
|
Jul 02, 2023 |
|
AF - Using (Uninterpretable) LLMs to Generate Interpretable AI Code by Joar Skalse
|
Jul 02, 2023 |
|
AF - Agency from a causal perspective by Tom Everitt
|
Jun 30, 2023 |
|
AF - When do "brains beat brawn" in Chess? An experiment by titotal
|
Jun 28, 2023 |
|
AF - Catastrophic Risks from AI #6: Discussion and FAQ by Dan H
|
Jun 27, 2023 |
|
AF - Catastrophic Risks from AI #5: Rogue AIs by Dan H
|
Jun 27, 2023 |
|
AF - Catastrophic Risks from AI #4: Organizational Risks by Dan H
|
Jun 26, 2023 |
|
AF - The fraught voyage of aligned novelty by Tsvi Benson-Tilsen
|
Jun 26, 2023 |
|
AF - Catastrophic Risks from AI #3: AI Race by Dan H
|
Jun 23, 2023 |
|
AF - Why Not Subagents? by johnswentworth
|
Jun 22, 2023 |
|
AF - An Overview of Catastrophic AI Risks #2 by Dan H
|
Jun 22, 2023 |
|
AF - An Overview of Catastrophic AI Risks #1 by Dan H
|
Jun 22, 2023 |
|
AF - The Hubinger lectures on AGI safety: an introductory lecture series by Evan Hubinger
|
Jun 22, 2023 |
|
AF - Causality: A Brief Introduction by Tom Everitt
|
Jun 20, 2023 |
|
AF - Ban development of unpredictable powerful models? by Alex Turner
|
Jun 20, 2023 |
|
AF - Mode collapse in RL may be fueled by the update equation by Alex Turner
|
Jun 19, 2023 |
|
AF - Experiments in Evaluating Steering Vectors by Gytis Daujotas
|
Jun 19, 2023 |
|
AF - Provisionality by Tsvi Benson-Tilsen
|
Jun 19, 2023 |
|
AF - Revising Drexler's CAIS model by Matthew Barnett
|
Jun 16, 2023 |
|
AF - [Replication] Conjecture's Sparse Coding in Small Transformers by Hoagy
|
Jun 16, 2023 |
|
AF - LLMs Sometimes Generate Purely Negatively-Reinforced Text by Fabien Roger
|
Jun 16, 2023 |
|
AF - MetaAI: less is less for alignment. by Cleo Nardo
|
Jun 13, 2023 |
|
AF - Virtual AI Safety Unconference (VAISU) by Linda Linsefors
|
Jun 13, 2023 |
|
AF - TASRA: A Taxonomy and Analysis of Societal-Scale Risks from AI by Andrew Critch
|
Jun 13, 2023 |
|
AF - Contingency: A Conceptual Tool from Evolutionary Biology for Alignment by clem acs
|
Jun 12, 2023 |
|
AF - ARC is hiring theoretical researchers by Paul Christiano
|
Jun 12, 2023 |
|
AF - Introduction to Towards Causal Foundations of Safe AGI by Tom Everitt
|
Jun 12, 2023 |
|
AF - Explicitness by Tsvi Benson-Tilsen
|
Jun 12, 2023 |
|
AF - Inference-Time Intervention: Eliciting Truthful Answers from a Language Model by likenneth
|
Jun 11, 2023 |
|
AF - How biosafety could inform AI standards by Olivia Jimenez
|
Jun 09, 2023 |
|
AF - Takeaways from the Mechanistic Interpretability Challenges by Stephen Casper
|
Jun 08, 2023 |
|
AF - What will GPT-2030 look like? by Jacob Steinhardt
|
Jun 07, 2023 |
|
AF - An Exercise to Build Intuitions on AGI Risk by Lauro Langosco
|
Jun 07, 2023 |
|
AF - A Playbook for AI Risk Reduction (focused on misaligned AI) by HoldenKarnofsky
|
Jun 06, 2023 |
|
AF - AISC end of program presentations by Linda Linsefors
|
Jun 06, 2023 |
|
AF - Algorithmic Improvement Is Probably Faster Than Scaling Now by johnswentworth
|
Jun 06, 2023 |
|
AF - Wildfire of strategicness by Tsvi Benson-Tilsen
|
Jun 05, 2023 |
|
AF - How to Think About Activation Patching by Neel Nanda
|
Jun 04, 2023 |
|
AF - Unfaithful Explanations in Chain-of-Thought Prompting by miles
|
Jun 03, 2023 |
|
AF - [Replication] Conjecture's Sparse Coding in Toy Models by Hoagy
|
Jun 02, 2023 |
|
AF - Think carefully before calling RL policies "agents" by Alex Turner
|
Jun 02, 2023 |