SPOT shows that state-of-the-art AI models detect fewer than one in five known errors in full scientific papers, with precision below 7%.
Multiplicative Chow-K\"unneth decomposition and homology splitting of configuration spaces
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
We construct a splitting of the cohomology of configuration spaces of points on a smooth proper variety with a multiplicative Chow--K\"unneth decomposition. Applied to hyperelliptic curves, this shows that the hyperelliptic Torelli group acts trivially on the rational cohomology of ordered configuration spaces of points. Moreover, if $H_{g,n}$ denotes the moduli space of $n$-pointed hyperelliptic curves, the Leray spectral sequence for the forgetful map $H_{g,n} \to H_g$ degenerates immediately, in sharp contrast to the forgetful map from $M_{g,n}$ to $M_g$. This allows for new detailed calculations of the cohomology of $M_{2,n}$ for $n \leq 5$, and the stable cohomology of $H_{g,n}$ for $n \leq 5$. We also give a detailed study of the cohomology of symplectic local systems on $M_2$.
citation-role summary
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
When AI Co-Scientists Fail: SPOT-a Benchmark for Automated Verification of Scientific Research
SPOT shows that state-of-the-art AI models detect fewer than one in five known errors in full scientific papers, with precision below 7%.