AI ALIGNMENT FORUM
AF

Wikitags

Formal Proof

Edited by plex, et al. last updated 26th Sep 2021

A Formal Proof is a finite sequence of steps from axiom(s) or previous derived proof(s) which strictly follow the allowed rules of inference of the in which it exists. They are used to establish statements as true within a mathematical framework in a way which can be independently verified with extremely high certainty, with the most reliable flavor of proof being machine-checked proofs generated by proof assistants since they have even less room for human error.

mathematical system
Subscribe
1
Subscribe
1
Discussion0
Discussion0
Posts tagged Formal Proof
46Compact Proofs of Model Performance via Mechanistic Interpretability
Lawrence Chan, rajashree, Adrià Garriga-Alonso, Jason Gross
1y
2
13AXRP Episode 40 - Jason Gross on Compact Proofs and Interpretability
DanielFilan
4mo
0
10Most Minds are Irrational
David Manheim
7mo
1
7Squeezing foundations research assistance out of formal logic narrow AI.
Donald Hobson
2y
0
52Davidad's Bold Plan for Alignment: An In-Depth Explanation
Charbel-Raphael Segerie, Gabin
2y
5
56A list of core AI safety problems and how I hope to solve them
davidad (David A. Dalrymple)
2y
12
49Limitations on Formal Verification for AI Safety
Andrew Dickson
11mo
1
30Towards Guaranteed Safe AI: A Framework for Ensuring Robust and Reliable AI Systems
Joar Skalse
1y
8
23Alignment proposals and complexity classes
Evan Hubinger
5y
26
8Measuring Nonlinear Feature Interactions in Sparse Crosscoders [Project Proposal]
Jason Gross, rajashree
6mo
0
10Weak HCH accesses EXP
Evan Hubinger
5y
0
6Infra-Domain Proofs 2
Diffractor
4y
0
5Infra-Domain proofs 1
Diffractor
4y
0
4Proofs Section 2.1 (Theorem 1, Lemmas)
Diffractor
5y
0
4Proofs Section 1.1 (Initial results to LF-duality)
Diffractor
5y
0
Load More (15/21)
Add Posts