Sajib Acharjee Dip

Sajib Acharjee Dip

He/him

Ph.D. researcher in LLMs, post-training, multimodal learning, and agentic AI.

Posts by Tags

computational biology

Context Is Part of the Signal: Two Studies in Computational Pathology and Functional Genomics

3 minute read

Published:

Modern biological machine learning increasingly relies on pretrained representations. A pathology foundation model can transform an image into rich embeddings, while a perturbation model can learn recurring transcriptional programs from large single-cell datasets. But representations alone do not always contain enough information to generalize reliably.

computational pathology

Context Is Part of the Signal: Two Studies in Computational Pathology and Functional Genomics

3 minute read

Published:

Modern biological machine learning increasingly relies on pretrained representations. A pathology foundation model can transform an image into rich embeddings, while a perturbation model can learn recurring transcriptional programs from large single-cell datasets. But representations alone do not always contain enough information to generalize reliably.

foundation models

Context Is Part of the Signal: Two Studies in Computational Pathology and Functional Genomics

3 minute read

Published:

Modern biological machine learning increasingly relies on pretrained representations. A pathology foundation model can transform an image into rich embeddings, while a perturbation model can learn recurring transcriptional programs from large single-cell datasets. But representations alone do not always contain enough information to generalize reliably.

functional genomics

Context Is Part of the Signal: Two Studies in Computational Pathology and Functional Genomics

3 minute read

Published:

Modern biological machine learning increasingly relies on pretrained representations. A pathology foundation model can transform an image into rich embeddings, while a perturbation model can learn recurring transcriptional programs from large single-cell datasets. But representations alone do not always contain enough information to generalize reliably.