Research Notes

Short distillations of papers I have an opinion about — what the work shows, and what I think it misses.

Language Models with Conformal Factuality Guarantees · Christopher Mohri, Tatsunori Hashimoto · ICML 2024 Jun 2026

Conformal guarantees are only as honest as the exchangeability they assume

This is the cleanest demonstration I know that you can wrap a language model's output in a distribution-free correctness guarantee instead of a vibe. What I keep flagging is the fine print. The guarantee rides on exchangeability, and the settings I actually care about, agents acting over time, multimodal inputs, free-form generation, low-resource languages, are exactly the ones where exchangeability quietly fails. I read this less as a solved problem and more as the strongest available foundation to build the non-exchangeable, multimodal version on. That gap is most of why I think guaranteed abstention is still wide open.

Read the paper →
Representation Engineering: A Top-Down Approach to AI Transparency · Andy Zou et al. · arXiv 2023 Jun 2026

We can steer a model toward honesty. We have only checked that in English

Representation engineering made the case that high-level properties like honesty and harmlessness live in directions you can read off and push on, and it is fast becoming part of the default safety toolkit. My worry is methodological. Almost every result is validated in English, and we keep treating the probes and steering vectors as if they were language-agnostic without checking. If an honesty direction transfers cleanly to a low-resource language, that is a strong and useful claim. If it fails silently, we have shipped a monitor that is blind exactly where human oversight is already thinnest. Either way someone has to test it, and that is squarely the work I want to do.

Read the paper →

Featured In

AI rules in SEA: the risks, the fines, what you need to know
Tech in Asia · The Business Times Jun 2026

AI rules in SEA: the risks, the fines, what you need to know

Featured coverage on AI regulation across Southeast Asia, examining the emerging rules, enforcement risks, and compliance requirements shaping the region's AI landscape.

Medium @feliren

Thoughts on life, philosophy, AI, research, and engineering from my Medium articles.

Engineering Jan 2026

Your model just killed someone's grandmother: Why every production ML needs conformal prediction

How conformal prediction gives you honest uncertainty quantification with mathematical guarantees that actually hold in production.

Read on Medium →
Research May 2026

We all think our map is the territory

On ethnocentrism as an invisible cultural default, why our own cultural frameworks feel like objective reality, and what it takes to build global teams that can genuinely hold multiple maps at once.

Read on Medium →
Research Nov 2025

My initial exploration on Trustworthy AI

First encounters with Trustworthy AI, bridging information theory and responsible machine learning.

Read on Medium →
Engineering May 2026

Quality and Reliability for AI Engineers

A practical guide to building quality and reliability into AI systems, covering evaluation frameworks, failure modes, and engineering practices that keep production AI trustworthy at scale.

Read on Medium →
Personal May 2026

The cost of becoming

A poem for the people who are tired of waiting for themselves.

Read on Medium →
Engineering May 2026

Four hours, one vendor, one airline

What a four-hour American Airlines booking outage reveals about the compounding costs of vendor dependency in modern infrastructure.

Read on Medium →
Personal May 2026

If everyone forgets, why be good

On why moral action still matters even when no one remembers, exploring Mother Teresa, Aristotle, and what we owe to the universe.

Read on Medium →
Personal Jan 2026

The art of letting go while still caring

What I learned from trying to control everything and ending up with nothing.

Read on Medium →
Personal Dec 2025

Meditation. 7 days. No phone. Noble silence.

A data-driven researcher's journey into 7 days of silent meditation at Taman Brahma Bali Usada.

Read on Medium →
Personal May 2025

Phoenix Protocol: Rising from the ashes of adversity

On resilience, reinvention, and learning to rise again after failure and loss.

Read on Medium →
Engineering Oct 2024

Privacy Matters: Navigating the Digital Age with Confidence

Discussing data synthesis, privacy-preserving analytics, and responsible data stewardship.

Read on Medium →
Personal Jul 2024

Mental Health Matters: Please Take Care

A personal reflection on mental health awareness, self-care, and breaking the silence.

Read on Medium →
Personal May 2024

A Stranger's Guide to the IU Concert

Reflections on fandom, culture, and music through the lens of an IU concert experience.

Read on Medium →
Personal Mar 2024

Gone from social media for 4 years. Now I am back

Returning to social media after 4 years, reflections on digital presence and personal growth.

Read on Medium →