eyes robson
LightNews LightNews
eyesrobson.bsky.social
eyes robson
@eyesrobson.bsky.social
89 followers 260 following 16 posts
PhD Candidate at UC Berkeley // y = mx + biology // bioethics, algorithmic fairness, and genomic AI // they/she 🏳️‍⚧️
Posts Replies Media Videos
eyesrobson.bsky.social
breaking: ExxonMobil announces "it's actually carbon-neutral" after adjusting for factors in the petroleum refinement process "outside of its control"
Screenshot from Andrew Ng's The Batch Newsletter, which reads "Gemini's Environmental Impact Measured" before stating Google's claim that its "environmental footprint [is smaller] than previous estimates had led it to expect. Not show is that the "research" ignored most of the actual necessary training & infrastructure to produce such models, which is known to increase the carbon emissions by two orders of magnitude (i.e. The Verge's 2023 analysis, which is mentioned briefly in The Batch's newsletter)
September 3, 2025 at 8:03 PM
eyesrobson.bsky.social
we also present an undervalued interpretability approach, which decomposes model variance explained into 'interpretable variables' like GC-content etc.

Borzoi and Enformer capture deeper features than the ones we test out, even surprisingly cryptic chromosomal features from sequence alone
A figure of four stacked bar plots. Each bar plot sums to the R2 variance explained by a (model x task) combination, for the models Enformer and Borzoi. The two GUANinE tasks are dnase-propensity (accessibility, left two) and cons30 (conservation, right two). The blue portion of each bar is uniquely captured variance from the deep models, while the orange portion of each bar is shared variance (i.e. these models have absorbed useful signal from these factors, which include chromosomal variables like chromosome size). Nearly invisible, save for the rightmost (Borzoi, cons30) task, is the green, interpretable variable (IV) only portion of each bar -- this means Enformer and Borzoi have extracted all of the useful signal, subsuming these features entirely, despite never having seen them during sequence-only training.
January 31, 2025 at 9:35 PM
eyesrobson.bsky.social
we also examined the accuracy of Borzoi at different bin averaging scales -- for the region-based tasks of GUANinE, more bins = better perf., aside from the accessibility task
A 3x2 table of six plots, each showing the L2-regularized few-shot performance of Borzoi on a different GUANinE task, with the x-axes showing different bin averaging sizes (corresponding to 32 to 416bp). Most plots show performance slightly up and to the right, except for the top-left plot for the dnase-propensity task (an accessibility task). Enformer's 3-bin performance is also present in each plot, usually below Borzoi's at the same x-axis resolution (384 bp), except for the two conservation tasks (center), where Enformer significantly outperforms Borzoi.
January 31, 2025 at 9:26 PM
eyesrobson.bsky.social
walkable city?? this code ain't walkable!!
R code of install.packages('car') -- a common statistical package... named 'car' ... yikes
February 15, 2024 at 10:51 PM