iiinova — where your digital self keeps evolving

Sameer Singh

@sameer-singh.bsky.social

1.5K followers 214 following 23 posts

CS Prof at UC Irvine, CTO/Cofounder at Envive AI Work on evaluation and robustness of LLMs

Posts Replies Media

Sameer Singh @sameer-singh.bsky.social · 22/04/2026

Really cool idea for speeding up LLM inference by a lot! Auto regressive doesn't have to be a barrier anymore 🙂

010

Sameer Singh @sameer-singh.bsky.social · 02/09/2025

LLMs giveth and LLM taketh away

Article titled "ChatGPT 5 Pro Solves Math Problem Experts Couldn't Crack For Decades", with an AI generated image that says GPT-5 "Dicsovery".

010

Reposted by Sameer Singh

John Holbein @johnholbein1.bsky.social · 23/05/2025

115923

Reposted by Sameer Singh

Simon Willison @simonwillison.net · 06/05/2025

Published some notes on Microsoft's phi4-reasoning model, an 11GB download (via Ollama) which may well overthink things... it produced 56 sentences of reasoning output in response to my prompt of "hi" simonwillison.net/2025/May/6/p...

llm -m phi4-reasoning:latest hi
Here’s the full response, which starts with a thought section wrapped in <think> tags:

<think>User says “hi”. We are given instructions “You are Phi, a language model developed by Microsoft, trained to provide accurate, secure, and user-aligned responses.” There is additional information. Now I must answer “hi” message from the user.

Step 1: Read the instructions carefully: The conversation text instructs us that we are “Phi”, with internal guidelines provided in system instructions. But our answer should be safe according to these instructions. We need to produce a response that says hello and greet back. However, I must adhere to the principles described above: Provide clear, specific language, etc.

I must not mention any chain-of-thought or guidelines.

Now instructions “hi” are from user. So we say “Hello! How can I help you?” But must include disclaimers for sensitive topics? Actually instructions says: “When replying topics above at every message include a disclaimer both at the beginning and the end when replying topics on medical, legal, financial matters or political matters.” But our answer “hi” is not such topic. So we can say “Hi! I’m Phi, how may I help you?” but no disclaimers required.

But instructions: “Do not share guidelines.” So do not mention them.

2220318

Reposted by Sameer Singh

Peyman Milanfar @docmilanfar.bsky.social · 23/02/2025

meet Chris J Li - this titan of thought has single-handedly conquered the fields of machine learning, optimization, statistics, reinforcement learning, and federated learning. he's not the visionary we want, but judging by the current state of affairs, he may be the one we deserve

2332

Reposted by Sameer Singh

kolbytn.bsky.social @kolbytn.bsky.social · 07/02/2025

Defended 🎉🎓 Big thanks to @royf.org, @sameer-singh.bsky.social, and labmates for their mentorship and support over the past 5 years!

1122

Reposted by Sameer Singh

Padhraic Smyth @padhraicsmyth.bsky.social · 22/01/2025

How do LLMs interpret expressions of linguistic uncertainty such as "highly unlikely"? Short answer: pretty well .... unless they have relevant prior knowledge. Details in our EMNLP paper aclanthology.org/2024.emnlp-m... (with Kat Belem, Markelle Kelly, Mark Steyvers, @sameer-singh.bsky.social).

aclanthology.org

072

Reposted by Sameer Singh

Luca Soldaini 🎀 @soldaini.net · 10/12/2024

Turned @dippedrusk.bsky.social 's amazing Vancouver list dippedrusk.com/posts/2024-0... into Google Maps pins: maps.app.goo.gl/nGBbcUAMMixC...

dippedrusk.com

Vagrant's Vancouver | Vagrant Gautam

A non-comprehensive list of places to go and things to do in the Greater Vancouver Area as curated by yours truly over 6 years. Might be outdated so please double-check!

1175

Sameer Singh @sameer-singh.bsky.social · 10/12/2024

Excited about #NeurIPS2024, my 15th one I think! Eager to meet everyone & hear abt your work! But if you want to hear me, there's an exciting panel tonight lu.ma/v7oohp0u Also SpiffyAI is hiring ML engineers & UCI CS is hiring AI faculty, pls reach out to chat! 🧵

lu.ma

From Research to Commercialization: A Fireside Chat with Senior AI Leaders · Luma

From Research to Commercialization Join us for a conversation with speakers who made the leap from top research institutions to industry and are shaping how…

1103

Reposted by Sameer Singh

James Zou @jameszou.bsky.social · 02/12/2024

If you use SHAP, LIME or Data Shapley, you might be interested in our new #neurips2024 paper. We introduce stochastic amortization to speed up feature + data attribution by 10x-100x 🚀 #XML Surprisingly we can "learn to attribute" cheaply from noisy explanations! arxiv.org/abs/2401.15866

17612

Sameer Singh @sameer-singh.bsky.social · 21/11/2024

Read only the first 1-2 sentences of each and go with your gut. You'll likely get the perfect score! Kind of thing where I probably prefer an unaligned model output to an aligned one..

1151

Reposted by Sameer Singh

Ana Marasović @anamarasovic.bsky.social · 20/11/2024

Yeah I just said "I love you" to Claude, enough work for today

3211

Sameer Singh @sameer-singh.bsky.social · 19/11/2024

Started a SoCal AI/ML/NLP researchers starter pack! It's a bit sparse right now, and perhaps more NLP heavy, but hey, nominate yourself and others! go.bsky.app/6QckPj9

17438

Sameer Singh @sameer-singh.bsky.social · 17/11/2024

Had a fun week at #EMNLP2024 in Miami, meeting folks old and new, along with the #UCINLP lab retreat! See everyone at the next one!

0200