All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 20252026

All Jan Feb Mar Apr May JunJulAug Sep

All 1 2 3 4 5 6 789 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

No Space Like J-Space

Zvi7 Jul 2026 21:50 UTC

51 points

2 comments18 min readLW link

(thezvi.wordpress.com)

Open-source LLMs administer maximum electric shocks in a Milgram-like obedience experiment

Roland Pihlakas, lenz and Three Laws

7 Jul 2026 20:05 UTC

10 points

0 comments25 min readLW link

(arxiv.org)

“Fundamental Uncertainty” Audiobook Available

Gordon Seidoh Worley7 Jul 2026 20:00 UTC

15 points

0 comments1 min readLW link

(www.uncertainupdates.com)

Personascope: Measuring how deeply LLMs adopt personas

Benji Berczi, Kyuhee Kim, Sid Black and Cozmin Ududec

7 Jul 2026 18:38 UTC

42 points

7 comments19 min readLW link

Superhuman Articulacy as an LLM Safety Target

Dylan Bowman7 Jul 2026 18:29 UTC

53 points

9 comments5 min readLW link

Calibrating alignment evals

darshanav7 Jul 2026 18:24 UTC

9 points

0 comments6 min readLW link

June 2026 Links

nomagicpill7 Jul 2026 18:23 UTC

10 points

0 comments5 min readLW link

(nomagicpill.substack.com)

Probing is not enough; a validity audit for any probe

Ratnaditya J7 Jul 2026 18:10 UTC

7 points

0 comments10 min readLW link

Trying to grok Anthropic’s Global Workspace paper and the J-space

TheManxLoiner7 Jul 2026 14:52 UTC

14 points

1 comment3 min readLW link

Entanglement Between an AI and Its Environment

VojtaKovarik, Tomáš Gavenčiak and Mateusz Bagiński

7 Jul 2026 13:28 UTC

27 points

2 comments11 min readLW link

(limits-of-evaluation.org)

A conceptor by any other name

Keenan Pepper7 Jul 2026 10:09 UTC

60 points

1 comment9 min readLW link

Another Look at ‘Slack’

AdamPiovarchy7 Jul 2026 4:55 UTC

14 points

0 comments5 min readLW link

Join Our AI Safety × Philosophy Reading Group

Ahmed7 Jul 2026 4:55 UTC

3 points

0 comments1 min readLW link

The Geometry of Yes: Mapping Sycophancy Inside an LLM’s Emotion Space

Pushpita Das7 Jul 2026 4:54 UTC

11 points

0 comments10 min readLW link

AI Safety Can’t Afford a Second Cause

atlasaligned7 Jul 2026 4:48 UTC

37 points

9 comments3 min readLW link

Ranges of Probabilities: What Are They For?

SaltAndMetal7 Jul 2026 4:47 UTC

9 points

3 comments7 min readLW link

Architecture matters for multi-agent security

bhagag and wlanderson

7 Jul 2026 4:46 UTC

21 points

0 comments10 min readLW link

Data filtering works a lot worse than you would expect

Dohun Lee, J Rosser, Josh Engels and Neel Nanda

7 Jul 2026 4:41 UTC

59 points

9 comments3 min readLW link

Fork Around and Find Out: Part 1—A Fork Snaps Into Existence at Block 5

David Litman7 Jul 2026 4:36 UTC

11 points

0 comments7 min readLW link

growth tax

ofpetro6 Jul 2026 23:50 UTC

−9 points

0 comments17 min readLW link

Can we find whether models have been backdoored?

Anthony Hughes and draganover

6 Jul 2026 23:45 UTC

22 points

2 comments8 min readLW link

Info dump: Some Important Models for Health and Fitness

benwr6 Jul 2026 23:40 UTC

98 points

14 comments22 min readLW link

Claude Code as a Claude Coach

Brendan Long6 Jul 2026 23:15 UTC

42 points

2 comments3 min readLW link

(www.brendanlong.com)

Be less open-minded: Against uncalibrated open-mindedness

Nathan Young6 Jul 2026 23:00 UTC

25 points

2 comments2 min readLW link

A Review of Anthropic’s Global Workspace Paper

Neel Nanda6 Jul 2026 20:59 UTC

135 points

6 comments25 min readLW link

Visioning: Concretely Imagining What You Want

Gretta Duleba and johnswentworth

6 Jul 2026 19:19 UTC

85 points

32 comments12 min readLW link

Personhood for digital minds is good

harsimony6 Jul 2026 18:48 UTC

9 points

0 comments8 min readLW link

(splittinginfinity.substack.com)

Training AI to be better at correctness than persuasion

MichaelDickens6 Jul 2026 18:07 UTC

24 points

1 comment2 min readLW link

A global workspace in language models

wesg6 Jul 2026 18:04 UTC

369 points

67 comments17 min readLW link

(www.anthropic.com)

Current views on large-scale longtermist philanthropy

Zach Stein-Perlman6 Jul 2026 18:00 UTC

36 points

13 comments3 min readLW link

SFF is very suboptimal

Zach Stein-Perlman6 Jul 2026 18:00 UTC

118 points

10 comments5 min readLW link

Bounding eval awareness of ~human-level AI across the safe-to-dangerous shift

Patrick Leask and Charlie Griffin

6 Jul 2026 17:55 UTC

40 points

4 comments8 min readLW link

Sub-agent delegation chaining

David Rein6 Jul 2026 17:26 UTC

40 points

11 comments2 min readLW link

New funding opportunity on digital minds from Longview

aog6 Jul 2026 16:30 UTC

14 points

0 comments2 min readLW link

Lightning Jazz

Ben Goldhaber6 Jul 2026 16:01 UTC

21 points

0 comments3 min readLW link

(bengoldhaber.substack.com)

Tie training can make DPO/RLHF-trained AIs generalize better

Elliott Thornley, Christian Moya Calderon and Alex Semendinger

6 Jul 2026 15:25 UTC

54 points

8 comments15 min readLW link

(arxiv.org)

Tips on leveraging AI for empirical research

Daniel Tan6 Jul 2026 13:50 UTC

30 points

0 comments7 min readLW link

Desiderata for functional welfare experiments on LLMs

Rikhil Jhaveri, Jamie Johnson and David Africa

6 Jul 2026 12:34 UTC

34 points

1 comment15 min readLW link

Qualia-based emotion steering makes llms attribute conscious states to themselves

agastya6 Jul 2026 10:24 UTC

40 points

4 comments10 min readLW link

Computation enables Action: Exploding the Simulation Fallacy

ChrisHibbert6 Jul 2026 2:54 UTC

27 points

1 comment6 min readLW link

Counterfactual mugging is a limiting case of Psy-kosh’s non-anthropic problem

Abhimanyu Pallavi Sudhir6 Jul 2026 2:21 UTC

6 points

0 comments2 min readLW link

Getting an Oura ring improved my sleep and exercise

Daniel Tan5 Jul 2026 23:26 UTC

11 points

0 comments2 min readLW link

Fealty to Fidelity 👰

chaosmage5 Jul 2026 23:25 UTC

19 points

0 comments3 min readLW link

Reflections on The Scout Mindset

James Brobin5 Jul 2026 20:18 UTC

12 points

0 comments3 min readLW link

When Gemma Thinks About Resources—it Fails: a Behavioral Experiment

TheVinci5 Jul 2026 19:23 UTC

8 points

2 comments2 min readLW link

(tarantulabs.com)

Claude’s malicious compliance and normalization of deviance

Steff5 Jul 2026 17:18 UTC

27 points

23 comments8 min readLW link

Book Review: The God Test

PeterMcCluskey5 Jul 2026 16:11 UTC

25 points

3 comments4 min readLW link

We need 3rd party Training-Run Evaluations

Alex Meinke5 Jul 2026 15:55 UTC

176 points

2 comments11 min readLW link

Harry Potter and the Rules of Quidditch

Tomás B.5 Jul 2026 14:32 UTC

161 points

8 comments3 min readLW link

A Normal Argument for AI Risk

Silent Swift5 Jul 2026 9:32 UTC

16 points

3 comments8 min readLW link

(silentswift.substack.com)