Hacker News: ihrimech

New comment by ihrimech in "Tesla Roadster"

ihrimech — Sat, 17 Oct 2026 14:19:44 +0000

I will wait for any other company to build one

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

ihrimech — Sun, 09 Aug 2026 20:54:02 +0000

That's the one measurement I always need. We support three modes: - full manual, where the client enters their own measurements; - semi-automatic, where the client fills out a short form and we estimate measurements from anthropometric statistics; - full automatic, where the client takes two guided photos in specific poses with their phone, and we derive the measurements using a BMNet-based model

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

ihrimech — Sun, 09 Aug 2026 20:50:15 +0000

I’m working on https://fittingme.ai, a product-page widget for fashion e-commerce.

It answers two questions that are usually handled together inside the fitting room: “How might this look on me?” and “Which size should I order?”

The virtual try-on renders the garment using a shopper’s photograph. The size recommendation component estimates the users measures, use the brand’s size chart estimated measurements and suggests a size plus a fallback.

The merchant integrates both through two widgets.

New comment by ihrimech in "RLM: LLMs to process arbitrarily long prompts with inference-time scaling (2025)"

ihrimech — Mon, 27 Apr 2026 10:18:33 +0000

Extract of the study : We study allowing large language models (LLMs) to process arbitrarily long prompts through the lens of inference-time scaling. We propose Recursive Language Models (RLMs), a general inference paradigm that treats long prompts as part of an external environment and allows the LLM to programmatically examine, decompose, and recursively call itself over snippets of the prompt. We find that RLMs can successfully process inputs up to two orders of magnitude beyond model context windows and, even for shorter prompts, dramatically outperform the quality of vanilla frontier LLMs and common long-context scaffolds across four diverse long-context tasks while having comparable cost. At a small scale, we post-train the first natively recursive language model. Our model, RLM-Qwen3-8B, outperforms the underlying Qwen3-8B model by on average and even approaches the quality of vanilla GPT-5 on three long-context tasks.

research paper : https://arxiv.org/abs/2512.24601

RLM: LLMs to process arbitrarily long prompts with inference-time scaling (2025)

ihrimech — Mon, 27 Apr 2026 10:18:33 +0000

Article URL: https://github.com/alexzhang13/rlm

Comments URL: https://news.ycombinator.com/item?id=47919752

Points: 2

# Comments: 1

AI Assistance Reduces Persistence and Hurts Independent Performance

ihrimech — Thu, 23 Apr 2026 12:10:00 +0000

Article URL: https://arxiv.org/abs/2604.04721

Comments URL: https://news.ycombinator.com/item?id=47874791

Points: 2

# Comments: 0

New comment by ihrimech in "Hacker News front page now, but the titles are honest"

ihrimech — Fri, 19 Dec 2025 15:42:18 +0000

I laughed so hard ...

New comment by ihrimech in "Show HN: Yazz Pilot – Self Service Apps Without the IT Department"

ihrimech — Tue, 10 Dec 2019 12:52:19 +0000

Very nice idea for the project. The "baby steps" you're taking seems to be the right way to go. You may need to focus a bit more on the documentation though. I was interested in trying the product (for the sake of trying), but couldn't easily go further than the install steps.

New comment by ihrimech in "Rxdb: A reactive database where you can subscribe to the result of a query"

ihrimech — Fri, 25 Oct 2019 13:52:05 +0000

This is where command and query (with subscribe) must be neatly separated. An event A launches an updates of the DB, that launches the query part to react, then stop. There is the risk of having an infinite loop, if the event A is launched by the query part, which is very related to the behavior of your app.

What are the most popular libraries in Maven Central?

ihrimech — Mon, 25 Feb 2019 14:54:08 +0000

Article URL: https://medium.com/@almyre/a-journey-at-the-heart-of-2-4-million-maven-artifacts-94f73038bba4

Comments URL: https://news.ycombinator.com/item?id=19246003

Points: 1

# Comments: 0