korrents

On the map

Tap a claim on the ring to put it at the centre.

← Information should be expected in a model's activations when it has a causal impact on the actions the model takes.

17 connected korrents · 10 moments from 10 Dec 2015 to 23 Aug 2026.

Everything filed under neural networks neural networks Everything filed under LLMs LLMs Everything filed under metascience metascience Same subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subject Read this korrent: Information should be expected in a model's activations when it has a causal impact on the actions the model takes. Information should be expected in amodel's activations when it has a causalimpact on the actions the model takes. Last stated 11 months ago 20 Dec 2025 NN Neel Nanda — holds since 2025-12-20 — tap for who they are Same subject: A model may internally represent that it is uncertain, but that signal is not wired up to it saying in words that it does not know. — tap to centre the map on it A model may internally representthat it is uncertain, but thatsignal is not wired up to it sayingin words that it does not know. Last stated 2 years ago 5 Feb 2025 AK Andrej Karpathy — holds since 2025-02-05 — tap for who they are Same subject: When models see an entity in a prompt, they often look up everything they know about it on its name, not only when needed. — tap to centre the map on it When models see an entity in aprompt, they often look upeverything they know about it on itsname, not only when needed. Last stated 11 months ago 20 Dec 2025 NN Neel Nanda — holds since 2025-12-20 — tap for who they are Same subject: Using a model as its own activation oracle introduces correlated risk if the model is misaligned. — tap to centre the map on it Using a model as its own activationoracle introduces correlated risk ifthe model is misaligned. Last stated 11 months ago 20 Dec 2025 NN Neel Nanda — holds since 2025-12-20 — tap for who they are Same subject: The more uncertain what follows a token, the more useful it is for a model to represent all relevant context on it. — tap to centre the map on it The more uncertain what follows atoken, the more useful it is for amodel to represent all relevantcontext on it. Last stated 11 months ago 20 Dec 2025 NN Neel Nanda — holds since 2025-12-20 — tap for who they are Same subject: There is little good evidence of anything achieved by analyzing mechanisms that could not have been done by analyzing activations. — tap to centre the map on it There is little good evidence ofanything achieved by analyzingmechanisms that could not have beendone by analyzing activations. Last stated 11 months ago 20 Dec 2025 NN Neel Nanda — holds since 2025-12-20 — tap for who they are Same subject: If you want the model to pay attention to a certain thing, you have to say that thing. — tap to centre the map on it If you want the model to payattention to a certain thing, youhave to say that thing. Last stated 7 months ago 14 Apr 2026 BL Brian Lovin — holds since 2026-04-14 — tap for who they are Same subject: That the brain produces the mind is still just an assumption; we have no working theory of how it happens — tap to centre the map on it That the brain produces the mind isstill just an assumption; we have noworking theory of how it happens Last stated 3 years ago 11 Jan 2024 YH Yuval Noah Harari — holds since 2024-01-11 — tap for who they are Same subject: The industry was right to focus on model intelligence, but has not yet thought enough about giving a model more context than any person could have so it can advise on big decisions. — tap to centre the map on it The industry was right to focus onmodel intelligence, but has not yetthought enough about giving a modelmore context than any person couldhave so it can advise on bigdecisions. Last stated 3 months ago 23 Aug 2026 SA Sam Altman — holds since 2026-08-23 — tap for who they are Same subject: A curve, which is what a deep learning model is, is a very bad substrate for discrete computation. — tap to centre the map on it A curve, which is what a deeplearning model is, is a very badsubstrate for discrete computation. Last stated 2 years ago 6 Nov 2024 FC François Chollet — holds since 2024-11-06 — tap for who they are Same subject: A layer should learn a residual with reference to its own input rather than an unreferenced function, which is what makes great depth trainable. — tap to centre the map on it A layer should learn a residual withreference to its own input ratherthan an unreferenced function, whichis what makes great depth trainable. Last stated 11 years ago 10 Dec 2015 JS Jian Sun — holds since 2015-12-10 — tap for who they are KH Kaiming He — holds since 2015-12-10 — tap for who they are Same subject: A project-level agent memory file works like a neural net, steering the agent much as a fine-tuned LoRA steers a model. — tap to centre the map on it A project-level agent memory fileworks like a neural net, steeringthe agent much as a fine-tuned LoRAsteers a model. Last stated 3 months ago 23 Aug 2026 KC Kun Chen — holds since 2026-08-23 — tap for who they are Same subject: A small enough winning ticket learns faster than the network it was cut out of, and ends up more accurate than it. — tap to centre the map on it A small enough winning ticket learnsfaster than the network it was cutout of, and ends up more accuratethan it. Last stated 9 years ago 9 Mar 2018 MC Michael Carbin — holds since 2018-03-09 — tap for who they are JF Jonathan Frankle — holds since 2018-03-09 — tap for who they are Same subject: A winning ticket wins on its initial weights: the connections it keeps started at values that happen to make training work. — tap to centre the map on it A winning ticket wins on its initialweights: the connections it keepsstarted at values that happen tomake training work. Last stated 9 years ago 9 Mar 2018 MC Michael Carbin — holds since 2018-03-09 — tap for who they are JF Jonathan Frankle — holds since 2018-03-09 — tap for who they are Same subject: Intelligence is the execution of small, simplicity-weighted programs found by search over smooth loss landscapes in overparameterized networks, approaching Bayes-optimal reasoning as scale increases. — tap to centre the map on it Intelligence is the execution ofsmall, simplicity-weighted programsfound by search over smooth losslandscapes in overparameterizednetworks, approaching Bayes-optimalreasoning as scale increases. Last stated 5 years ago 11 Jun 2021 GW Gwern — holds since 2021-06-11 — tap for who they are Same subject: A transduction model can rely entirely on self-attention for input and output representations without sequence-aligned RNNs or convolution. — tap to centre the map on it A transduction model can relyentirely on self-attention for inputand output representations withoutsequence-aligned RNNs orconvolution. Last stated 9 years ago 12 Jun 2017 IP Illia Polosukhin — holds since 2017-06-12 — tap for who they are JU Jakob Uszkoreit — holds since 2017-06-12 — tap for who they are NP Niki Parmar — holds since 2017-06-12 — tap for who they are AV Ashish Vaswani — holds since 2017-06-12 — tap for who they are NS Noam Shazeer — holds since 2017-06-12 — tap for who they are AG Aidan N. Gomez — holds since 2017-06-12 — tap for who they are +2 2 more Same subject: Linear probes are very strong classifiers, and attention-head probes are even stronger. — tap to centre the map on it Linear probes are very strongclassifiers, and attention-headprobes are even stronger. Last stated 11 months ago 20 Dec 2025 NN Neel Nanda — holds since 2025-12-20 — tap for who they are Same subject: Self-attention can yield more interpretable models. — tap to centre the map on it Self-attention can yield moreinterpretable models. Last stated 9 years ago 12 Jun 2017 IP Illia Polosukhin — holds since 2017-06-12 — tap for who they are JU Jakob Uszkoreit — holds since 2017-06-12 — tap for who they are NP Niki Parmar — holds since 2017-06-12 — tap for who they are AV Ashish Vaswani — holds since 2017-06-12 — tap for who they are NS Noam Shazeer — holds since 2017-06-12 — tap for who they are AG Aidan N. Gomez — holds since 2017-06-12 — tap for who they are +2 2 more
same subject or similar wordinga cloud: claims about one subject, named for itbar: when it was last stated, on a scale from 2015 to today — full is todaya face: someone who holds the claim — tap it for who they are

At the centre Information should be expected in a model's activations when it has a causal impact on the actions the model takes. Last stated 20 Dec 2025 · 11 months ago Holds Neel Nanda Read this korrent →