---
title: "Abish Kamran (@AbishKamran) | Hashnode"
description: "Web Developer | AI/ML Enthusiast. Read the latest articles by Abish Kamran on Hashnode."
url: "https://hashnode.com/@AbishKamran"
---

@AbishKamran

# Abish Kamran

@AbishKamran · Joined January 2023

Web Developer | AI/ML Enthusiast

### [Agent Skills Explained: Teaching Your AI Agent How to Think.](https://abishkamran.hashnode.dev/agent-skills-explained-teaching-your-ai-agent-how-to-think?utm_source=hashnode&utm_medium=feed)

Learn what AI agent skills are, how SKILL.md works, and how to build smarter, efficient AI agents using structured instructions.

6 min read · [abishkamran.hashnode.dev](https://abishkamran.hashnode.dev/agent-skills-explained-teaching-your-ai-agent-how-to-think?utm_source=hashnode&utm_medium=feed)

[Apr 1](https://hashnode.com/posts/agent-skills-explained-teaching-your-ai-agent-how-to-think/69cd6c083085402b9c508e00)

### [Deeper Transformers Are Forgetting What They Learned. MoDA Is the Fix.](https://abishkamran.hashnode.dev/deeper-transformers-are-forgetting-what-they-learned-moda-is-the-fix?utm_source=hashnode&utm_medium=feed)

Deep transformers forget early-layer signals. MoDA fixes this with depth KV caching — better performance, only 3.7% extra compute. Here's how it works

18 min read · [abishkamran.hashnode.dev](https://abishkamran.hashnode.dev/deeper-transformers-are-forgetting-what-they-learned-moda-is-the-fix?utm_source=hashnode&utm_medium=feed)

[Mar 25](https://hashnode.com/posts/deeper-transformers-are-forgetting-what-they-learned-moda-is-the-fix/69c41f5610e664c5dacf4a82)

### [Your Language Model Has Feelings About Your Text — And It Shows Up in the Attention Geometry](https://abishkamran.hashnode.dev/your-language-model-has-feelings-about-your-text-and-it-shows-up-in-the-attention-geometry?utm_source=hashnode&utm_medium=feed)

Papers I'm Reading — Issue #02 Paper: Emotion is Not Just a Label: Latent Emotional Factors in LLM Processing arXiv: 2603.09205 | cs.CL / cs.AI / cs.LG Authors: Benjamin Reichman, Adar Avasian, Samu

16 min read · [abishkamran.hashnode.dev](https://abishkamran.hashnode.dev/your-language-model-has-feelings-about-your-text-and-it-shows-up-in-the-attention-geometry?utm_source=hashnode&utm_medium=feed)

[Mar 13](https://hashnode.com/posts/your-language-model-has-feelings-about-your-text-and-it-shows-up-in-the-attention-geometry/69b3e23b4c737221668fe785)

### [Quantum Physics Just Showed Up in a Language Model. Here's What's Actually Going On.](https://abishkamran.hashnode.dev/quantum-physics-just-showed-up-in-a-language-model-here-s-what-s-actually-going-on?utm_source=hashnode&utm_medium=feed)

A deep dive into QISA — the paper that replaced dot-product attention with quantum fidelity inside GPT-1. What it does, how it works.

10 min read · [abishkamran.hashnode.dev](https://abishkamran.hashnode.dev/quantum-physics-just-showed-up-in-a-language-model-here-s-what-s-actually-going-on?utm_source=hashnode&utm_medium=feed)

[Mar 10](https://hashnode.com/posts/quantum-physics-just-showed-up-in-a-language-model-here-s-what-s-actually-going-on/69b00293abc0d9500156ebc9)

### [Designing a Scalable UI Automation Setup with Playwright](https://abishkamran.hashnode.dev/designing-a-scalable-ui-automation-setup-with-playwright-part-1-tech-stack-overview?utm_source=hashnode&utm_medium=feed)

Explore the process of designing a scalable user interface (UI) automation framework utilizing Playwright, Jest, and Allure to extract, validate, and report

3 min read · [abishkamran.hashnode.dev](https://abishkamran.hashnode.dev/designing-a-scalable-ui-automation-setup-with-playwright-part-1-tech-stack-overview?utm_source=hashnode&utm_medium=feed)

[Jan 6](https://hashnode.com/posts/designing-a-scalable-ui-automation-setup-with-playwright-part-1-tech-stack-overview/695d0bca77d504f8b8263738)