Hacker News: abhgh

New comment by abhgh in "Three sites made 215,128 “best software” pages for AI. Perplexity cites them"

abhgh — Thu, 03 Sep 2026 05:43:51 +0000

The searchable term is "Self-preference Bias" :)

My first encounter with any kind of study was the G-Eval paper [1]. They study whether their LLM judge prefers human or LLM-generated summaries (answer: it's the latter).

[1] Section 4 in https://aclanthology.org/2023.emnlp-main.153/

New comment by abhgh in "I accidentally turned LLM memory into program analysis"

abhgh — Sat, 29 Aug 2026 14:06:22 +0000

What you call "Weathering" has been a constant gripe of mine. We have LLM-driven softwares toward that almost seem to start from scratch every time a request comes in - there are mechanisms to learn or generalize, like writing out a memory, but they are not reliable or reliable in general. There is no convenient lever to be able to say "yes this is in the memory but the request seems like it needs a fresh scan of data, so ignore your memory", or the opposite "you can infer this from stuff in the memory - don't re-reason!". There is some work like Dynamic Cheatsheets [1] and Agentic Context Engineering [2] that have studied this aspect, but we are far from a generally reliable solution. And till we have that, I think the system variations for systems trying to solve this problem are going to be (a) LLM-leaning: create unstructured memory files, with human in the loop as a filter to reject inaccurate responses (b) LLM-as-a-layer: what you describe and the article kind of is doing.

[1] https://aclanthology.org/2026.eacl-long.333/ [2] https://openreview.net/pdf?id=eC4ygDs02R

New comment by abhgh in "Ternlight – 7 MB embedding model that runs in browser (WASM)"

abhgh — Tue, 07 Jul 2026 07:21:28 +0000

Cool project! I tried something similar a while ago [1] - I wanted to load up an embedding model and semantically order texts, all in the browser.

So I pull ONNX weights from HuggingFace (MPNet, MiniLM), use Transformers.js to embed, and use a clusterer from scikit-learn (running on pyiodide - it was a surprise to me that this worked flawlessly) on the page - all client-side.

[1] http://sol.quipu-strands.com/

New comment by abhgh in "Knoppix"

abhgh — Tue, 30 Jun 2026 23:49:03 +0000

Oh wow, they are still around! I used to keep a Knoppix disk around for times when my desktop would run out of space and wouldn't boot (if I remember right). Boot it with Knoppix, delete some files, and then boot it normally!

New comment by abhgh in "GLM 5.2 vs. Opus"

abhgh — Mon, 22 Jun 2026 14:15:15 +0000

You usually see this kind of analyses in conference papers, esp. if they have a datasets track. The NeurIPS Datasets & Benchmarks (D&B) track is a good example. But you will have to monitor the proceedings yourself closely - there is little chance of being accidentally exposed to them, because most blogs, announcements and popular media only mention a handful of the popular ones, e.g., Tau^2. For ex., across the years 2022, 2023 and 2024, 900+ papers were accepted in the D&B track [1] - of course, not all of them are LLM-related. I find them interesting because they often focus on specific system behaviors, and like you said, study them scientifically, so you can draw authoritative conclusions (or at least know specifically what part of a model's behavior you now know about, and what parts you don't).

[1] https://blog.neurips.cc/2025/09/30/reflecting-on-the-2025-re...

New comment by abhgh in "When I reject AI code even if it works"

abhgh — Sun, 21 Jun 2026 03:17:27 +0000

These "You're right to push back" scenarios are scary for me. I mostly code ML implementations, and some of the errors Claude Code (CC - have only used Opus 4.7) makes are very sneaky, and if you don't have sufficient experience in the area (I see this with people entering ML and writing their implementations with CC), you wouldn't know when to question CC and will let errors or future pitfalls silently slip into your code. A recent example was when there was data leakage in a model calibration step, which it refused to see as an error, till I wrote a detailed reason, and then it agreed that there was a "subtle leakage".

New comment by abhgh in "LLMs are eroding my software engineering career and I don't know what to do"

abhgh — Sun, 07 Jun 2026 14:37:55 +0000

Reg PRs - for the ones with complex requirements what I am seeing is that time to initial PR is very short, and a ping-pong between the reviewer and developer begins, because in my cases (not all) the developer vibe-coded parts, and they didn't really understand the requirements deeply or their code, and it takes multiple iterations for them to fix it. You can argue this is a human problem but this is the net effect I'm seeing.

I am not sure but for complex cases it seems to me that the earlier sum of moderately long PR time + moderately long review time has been replaced by very short PR time + even longer review time. I am not sure if there's a net gain in these cases. Sometimes even if the code is functionally correct, it's verbose enough (e.g., too many intermediate functions) that I think they will impact future reviews.

New comment by abhgh in "No more JetBrains products for me"

abhgh — Mon, 18 May 2026 23:33:43 +0000

I should go back to this (like some of the other comments suggest) - I think there is potential in suggesting multiple lines, and I would have really loved it to work, but it is clear that their ux testing was poor.

New comment by abhgh in "No more JetBrains products for me"

abhgh — Mon, 18 May 2026 21:20:25 +0000

I am not particularly bothered by the speed but the AI suggestion clutter is quickly becoming an issue for me :( I type fragments of a line and it will suggest the next 5-6 lines. They may not be outright wrong, but they might not represent the way I like to do things. Pressing Esc. and refocusing on what I was going to type in anyway is a disruptive experience.

New comment by abhgh in "Qwen3.5 Fine-Tuning Guide"

abhgh — Wed, 04 Mar 2026 15:52:39 +0000

They are great for specialized use-cases: (a) where the problem is not hard enough (you don't need reasoning), or (b) diverse enough (you don't need a world model), (c) you want cheap inference (and you can make it happen hardware-wise) and (d) you either have enough data or a workflow that accumulates data (with fine tuning with enough data you can sometimes beat a premier model while ensuring low latency - ofc, assuming (a) and (b) apply).

I make it sound like a rare perfect storm needs to exist to justify fine tuning, but these circumstances are not uncommon - to an extent (a), (c) and (d) were already prerequisites for deploying traditional ML systems.

New comment by abhgh in "Ask HN: What are you working on? (February 2026)"

abhgh — Mon, 09 Feb 2026 12:22:42 +0000

I notice you mentioned dspy - do you also support prompt optimization?

New comment by abhgh in "I miss thinking hard"

abhgh — Wed, 04 Feb 2026 07:13:44 +0000

This is an amazing quote - thank you. This is also my argument for why I can't use LLMs for writing (proofreading is OK) - what I write is not produced as a side-effect of thinking through a problem, writing is how I think through a problem.

The Gumbel-Max Trick

abhgh — Tue, 03 Feb 2026 17:32:03 +0000

Article URL: https://blog.quipu-strands.com/gumbel

Comments URL: https://news.ycombinator.com/item?id=46874077

Points: 2

# Comments: 0

New comment by abhgh in "Ask HN: When has a "dumb" solution beaten a sophisticated one for you?"

abhgh — Sun, 18 Jan 2026 07:37:56 +0000

I once modeled user journeys on a website using fancy ML models that honored sequence information, i.e., order of page visits, only to be beaten by bag-of-words (i.e., page url becomes a vector dimension, but order is lost) decision tree model, which was supposed to be my baseline.

What I had overlooked was that journeys on that particular website were fairly constrained by design, i.e., if you landed on the home page, did a bunch of stuff, put product X in the cart - there was pretty much one sequence of pages (or in the worst case, a small handful) that you'd traverse for the journey. Which means the bag-of-words (BoW) representation was more or less as expressive as the sequence model; certain pages showing up in the BoW vector corresponded to a single sequence (mostly). But the DT could learn faster with less data.

New comment by abhgh in "Vibe Coding Killed Cursor"

abhgh — Fri, 02 Jan 2026 17:12:52 +0000

I use Claude Code within Pycharm and I see the git diff format for changes there.

EDIT: It shows the side-by-side view by default, but it is easy to toggle to a unified view. There's probably a way to permanently set this somewhere.

New comment by abhgh in "A linear-time alternative for Dimensionality Reduction and fast visualisation"

abhgh — Tue, 16 Dec 2025 13:12:05 +0000

Thank you. Your comment about LLMs to semantically parse diverse data, as a first step, makes sense. In fact come to think of it, in the area of prompt optimization too - such as MIPROv2 [1] - the LLM is used to create initial prompt guesses based on its understanding of data. And I agree that UMAP still works well out of the box and has been pretty much like this since its introduction.

[1] Section C.1 in the Appendix here https://arxiv.org/pdf/2406.11695

New comment by abhgh in "A linear-time alternative for Dimensionality Reduction and fast visualisation"

abhgh — Tue, 16 Dec 2025 10:47:21 +0000

I was not aware this existed and it looks cool! I am definitely going to take out some time to explore it further.

I have a couple of questions for now: (1) I am confused by your last sentence. It seems you're saying embeddings are a substitute for clustering. My understanding is that you usually apply a clustering algorithm over embeddings - good embeddings just ensure that the grouping produced by the clustering algo "makes sense".

(2) Have you tried PaCMAP? I found it to produce high quality and quick results when I tried it. Haven't tried it in a while though - and I vaguely remember that it won't install properly on my machine (a Mac) the last time I had reached out for it. Their group has some new stuff coming out too (on the linked page).

[1] https://github.com/YingfanWang/PaCMAP

New comment by abhgh in "Algorithms for Optimization [pdf]"

abhgh — Mon, 01 Dec 2025 07:11:39 +0000

Thanks for the example. Yes, true, this is for expensive functions - to be precise functions that depend on data that is hard to gather, so you interleave the process of computing the value of the function with gathering strategically just as much data as is needed to compute the function value. The video on their page [1] is quite illustrative: calculate shortest path on a graph where the edge weights are expensive to obtain. Note how the edge weights they end up obtaining forms a narrow band around the shortest path they find.

[1] https://willieneis.github.io/bax-website/

New comment by abhgh in "Algorithms for Optimization [pdf]"

abhgh — Mon, 01 Dec 2025 05:40:42 +0000

Timefold looks very interesting. This might be irrelevant but have you looked at stuff like InfoBax [1]?

[1] https://willieneis.github.io/bax-website/

New comment by abhgh in "Terence Tao: At the Erdos problem website, AI assistance now becoming routine"

abhgh — Tue, 25 Nov 2025 17:35:19 +0000

You don't - the way I use LLMs for explanations is that I keep going back and forth between the LLM explanation and Google search /Wikipedia. And of course asking the LLM to cite sources helps.

This might sound cumbersome but without the LLM I wouldn't have (1) known what to search for, in a way (2) that lets me incrementally build a mental model. So it's a net win for me. The only gap I see is coverage/recall: when asked for different techniques to accomplish something, the LLM might miss some techniques - and what is missed depends upon the specific LLM. My solution here is asking multiple LLMs and going back to Google search.