---
title: "Binesh Kumar (@binesh) | Hashnode"
description: "Binesh Kumar (@binesh) writes on Hashnode. Read their latest articles and follow their work."
url: "https://hashnode.com/@binesh"
---

@binesh

# Binesh Kumar

@binesh · CT · Joined January 2026

### [Circuit Tracing: Finding Medical Features in Gemma 3](https://thedatasense.com/circuit-tracing-finding-medical-features-in-gemma-3?utm_source=hashnode&utm_medium=feed)

Gemma 3 uses Neuronpedia's tool to organize and activate medical knowledge contextually with prompts

4 min read · [thedatasense.com](https://thedatasense.com/circuit-tracing-finding-medical-features-in-gemma-3?utm_source=hashnode&utm_medium=feed)

[Feb 4](https://hashnode.com/posts/circuit-tracing-finding-medical-features-in-gemma-3/6983c33eb1ed7e5e35b2efca)

### [What Does Medical VLM Actually See? Experiments with MedGemma and Sparse Autoencoders](https://thedatasense.com/what-does-medical-vlm-actually-see-experiments-with-medgemma-and-sparse-autoencoders?utm_source=hashnode&utm_medium=feed)

Sparse Autoencoders improve medical image interpretability in MedGemma, affecting model responses based on question phrasing

5 min read · [thedatasense.com](https://thedatasense.com/what-does-medical-vlm-actually-see-experiments-with-medgemma-and-sparse-autoencoders?utm_source=hashnode&utm_medium=feed)

[Jan 24](https://hashnode.com/posts/what-does-medical-vlm-actually-see-experiments-with-medgemma-and-sparse-autoencoders/6976ccce5ff8814dbb117833)

### [Fluent But Wrong: LLM and Healthcare](https://thedatasense.com/fluent-but-wrong-llm-and-healthcare?utm_source=hashnode&utm_medium=feed)

Discusses limitations of deploying language models in healthcare, stressing the need for human verification and system improvements for safety

9 min read · [thedatasense.com](https://thedatasense.com/fluent-but-wrong-llm-and-healthcare?utm_source=hashnode&utm_medium=feed)

[Jan 20](https://hashnode.com/posts/fluent-but-wrong-llm-and-healthcare/697076b78b5a61c8ebc51d1a)

### [Opening the Black Box: How to See What Your Vision Language Model is Actually Looking At](https://thedatasense.com/opening-the-black-box-how-to-see-what-your-vision-language-model-is-actually-looking-at?utm_source=hashnode&utm_medium=feed)

Use Chefer et al.'s techniques to visualize model attention, enhancing trust in AI diagnostics through vision-language model explainability

19 min read · [thedatasense.com](https://thedatasense.com/opening-the-black-box-how-to-see-what-your-vision-language-model-is-actually-looking-at?utm_source=hashnode&utm_medium=feed)

[Jan 16](https://hashnode.com/posts/opening-the-black-box-how-to-see-what-your-vision-language-model-is-actually-looking-at/696d7c54f46e1acc28868b01)

### [Data Generating Process](https://thedatasense.com/data-generating-process?utm_source=hashnode&utm_medium=feed)

Understand the Data Generating Process (DGP) to explore mechanisms behind data, enabling better insight into uncertainty and statistical methods

7 min read · [thedatasense.com](https://thedatasense.com/data-generating-process?utm_source=hashnode&utm_medium=feed)

[Jan 14](https://hashnode.com/posts/data-generating-process/696ebff2745ecd51cbc4375a)
