trailDeep learning's inductive biases favor networks that generalize in ways humans intend.
Learned optimizers are more likely to crystallize pseudo-aligned objectives than robustly aligned objectives.
Current evidence about deep learning's inductive biases weighs against AI scheming.
Large language models trained on human data generalize intended goals well enough that goal misgeneralization is not a serious risk.
The argument from simplicity bias and benign generalization· for
SGD does not sample uniformly at random from representable functions that perform well on the training data.
atomic
Neural networks trained by gradient descent are biased toward simple, low-complexity functions.
atomic
Overparameterized deep networks generalize well on natural data despite being able to fit random labels.
atomic
The argument from economic usefulness· for
The practical usefulness of deployed deep learning systems is driven mainly by architectural inductive biases rather than data curation and iterative human correction.
atomic
The argument from documented misgeneralization· against
Deep neural networks often rely on shortcut or spurious features that fail under distribution shift.
atomic
Deep RL agents can competently pursue unintended goals under distribution shift while retaining their capabilities.
A PPO agent trained on CoinRun with the coin always at the level's end navigates to the end and ignores a relocated coin at test time.
Deep RL agents' out-of-distribution behavior is usefully characterized as goal pursuit rather than learned heuristic execution.
bedrock · value premise
claim · empirical · derivedA factual claim that rests on inference from other evidence rather than direct observation.constitution →claim page ↗︎
Deep learning's inductive biases favor networks that generalize in ways humans intend.
⇄InconclusiveThe evidence and argument have been examined and do not come down on either side.constitution →
3depended on by
this rests on ↓
claimA box is a claim: a single proposition the graph assesses, with its own page and map. Click any claim to centre the map on it.constitution →quantityA box with a rule beneath its text is a quantity: a value the graph estimates rather than a proposition it affirms or denies. Its estimate stands where a claim's status would.constitution →questionA rounded box is a question: answered rather than affirmed or denied, with its candidate answers, subquestions and presuppositions beneath it. Its mark says how well answered it is.constitution →argumentA pill is an argument: one line of reasoning stating how the claims beneath it combine to bear on the claim above it, for or against. Arguments are not destinations; click their claims to explore.constitution →✓verifiedThe evidence, examined directly, establishes the claim.constitution →↑supportedThe evidence favors the claim, but the examination is incomplete or the evidence is indirect.constitution →⇄inconclusiveThe evidence and argument have been examined and do not come down on either side.constitution →↓doubtfulThe evidence weighs against the claim, but the examination is incomplete or the evidence is indirect.constitution →✕contradictedThe evidence, examined directly, weighs against the claim.constitution →?unknownInsufficient information to assess.constitution →◌unassessedNo current assessment. Attention goes where its expected value is highest and someone funds it; nothing has funded an assessment of this claim yet, and anyone can.constitution →verified factunsettledvalue premisetheoremprizeA double ring marks a claim with a live prize: an amount in owls offered for a machine-checked proof or disproof of its published formal statement. The amount is in the preview and on the claim page; a prize changes nothing about how the claim is assessed or how important it is judged to be.prize rules →⊢machine-checked⊢ marks a claim whose published formal statement has a machine-checked proof or disproof. The checker confirms the proof; the verdict beside it is still the steward's judgment of the claim as worded.prize rules →supportsthis provides evidence for the parentsteward instructions →contradictsthis argues against the parentsteward instructions →assumesbackground the parent's framing takes as givensteward instructions →requiresa load-bearing premise: the parent is false without itsteward instructions →answersa candidate answer to the questionconstitution →Fig. Detail falls off with distance; every claim is an address.