Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
vikashjohn2505
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
2 ms
·
1.
▲
Beyond Solver Verdicts: Generative Reward Models for Autoformalizations
(arxiv.org)
2 points
by
vikashjohn2505
2mo ago
|
0 comments
2.
▲
Openleaf: Overleaf's workflow. Obsidian's interface. None of the cloud
(github.com)
1 points
by
vikashjohn2505
3mo ago
|
0 comments
3.
▲
Verge: Formal Refinement and Guidance Engine for Verifiable LLM Reasoning
(arxiv.org)
2 points
by
vikashjohn2505
10mo ago
|
2 comments
4.
▲
by
vikashjohn2505
10mo ago
We present a neurosymbolic framework that combines LLMs with SMT solvers to produce verification-guided answers through iterative refinement. Our approach decomposes LLM outputs into atomic claims, autoformalizes them into first-order logic