HNHacker News
TopNewBestAskShowJobs

napsternxg

304 karma · joined November 4, 2014

submissionscomments
napsternxg··on Show HN: Aiopandas – Async .apply() and .map() for Pandas, Faster API/LLMs Calls
This is a very clean api and I really like the way you implemented it directly in Pandas. I worked on something similar 2 years back but the API was not as this one. Thanks a lot to making this.
napsternxg··on Coping with dumb LLMs using classic ML
We often ignore the importance of using good baseline systems and jump to the latest shiny thing.

I had a similar experience few years back when participating in a ML competitions [1,2] for detecting and typing phrases in a text. I submitted an approach based on Named Enttiy Recognition using Conditional Random Field (CRF) which has been quite robust and well known in the community and my solution beat most of tuned Deep learning solutions by quite a large margin [1].

I think a lot of folks underestimate the complexity of using some of these models (DL, LLM) and just throw them at the problem or don't compare it well against well established baselines.

[1] https://scholar.google.com/citations?view_op=view_citation&h... [2] https://scholar.google.com/citations?view_op=view_citation&h...

napsternxg··on Evaluating the world model implicit in a generative model
We also did something similar in our NTULM paper at Twitter https://youtu.be/BjAmQjs0sZk?si=PBQyEGBx1MSkeUpX

Used in non generative language models like BERT but should help with generative models as well.

napsternxg··on Show HN: Probabilistic Tic-Tac-Toe
Nice game, I like the idea of your, opponent and no turn. I made a similar game long time back (around 2014) also called Probabilistic Tic Tac Toe ( https://shubhanshu.com/PT3/ ), but my randomization rules were different. I used coin toss to decide on the move vetween two choices.

Old HN thread: https://news.ycombinator.com/item?id=12932183

napsternxg··on Grounded language acquisition through the eyes and ears of a single child
Twitter thread from the author https://twitter.com/wkvong/status/1753132293491708027
napsternxg··on Try Redis in Browser
@mamato you you are right I made a typo in the url. It should be https://try.redis.io

Sorry about that. I can't delete this post hence posting this comment.

napsternxg··on Square to no longer return processing fees when buyers are issued a refund
Why cant US have something like UPI [1], which significantly reduced reliance of credit card companies and extra payment traaction fees in India. Is there some regulation holding this back or no state or federal government has tried it yet.

[1] https://en.wikipedia.org/wiki/Unified_Payments_Interface

napsternxg··on Barnes and Noble's surprising turnaround
Is it possible that the pandemic made people fall back in love with non digital assets as they were already exhausted by the excess screen use? Has that possibility been ruled out clearly?

I recently visited B&N and the feeling of moving around and sitting with books after spending all week in front of screen was quite refreshing.

napsternxg··on Probabilistic Tic Tac Toe game – Randomness adds challenge and trumps deadlocks
The motivation for the game as listed on the page, was to make the original tic-tac-toe (a timepass game), slightly interesting.

The coin toss forces the suboptimal move, thereby giving the other player a chance to play the optimal move next time. The original game is quite easy, and becomes boring because most games tend to be tied. This configurations will allow individuals to keep playing, for a longer duration, knowing that the other player might not be able to play the most optimal move.

napsternxg··on Tweet Sentiment analysis prediction, incremental training, visualization
This is a first GUI based tool for sentiment prediction of tweets using a model trained on research level tweet sentiment data. The tool allows a user to incrementally update the model with more information. It also supports a sentiment based timeline visualization of tweets.

All kind of feedback in welcome. Do contribute to the repository if you like.

napsternxg··on Visualizing Facebook Group Communications
This is a visualization to see communication patterns for facebook groups. The visualization is based on a class data of CS467 class at the University of Illinois at Urbana Champaign.