claim
active
claim:the-upper-bound-of-what-can-be-learned-from-a-dataset-is-not-the-most-capable-trajectory-but-the-conditional-structure-of-the-universe-implicated-by-their-sumThe upper bound of what can be learned from a dataset is not the most capable trajectory, but the conditional structure of the universe implicated by their sum.
Key insight about predictive learning's potential.
Source paper
extracted_fromNeighborhood — ranked by edge-count
Concepts (1)
concept
- The ideal limit of self-supervised learning as modeling the true conditional distribution.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Clarifies what unsupervised learning does.
- Janus's mathematical claim about exponential path combinatorics in transformers.
- Reframes the traditional life/mind hierarchy, supporting panpsychist implications.
- Motivating question.
- Unsupervised bottom-up tools are valuable because hypothesis generation is the hardest step.claim0.768Research philosophy.
- One of the updates about prosaic ML simulation.
- Extrapolation of scaling predictive models to AGI.
- Describes scaffolding method and the model's meta-learning loop.