Hacker News: 4b11b4

New comment by 4b11b4 in "The Emergent Symbolic Structure of Artificial Neural Networks"

4b11b4 — Wed, 02 Sep 2026 05:33:45 +0000

Sounds reasonable... That the model is sometimes learning a lossy vector representation of something symbolic in nature... Sure, a NN can approximate a function?

They say this holds in... Some examples they found?

I don't enough about this area

New comment by 4b11b4 in "I accidentally turned LLM memory into program analysis"

4b11b4 — Sat, 29 Aug 2026 14:23:42 +0000

Hmm DITA..

New comment by 4b11b4 in "Show HN: ThoughtDAG – An editable context graph for LLM conversations"

4b11b4 — Sat, 15 Aug 2026 14:52:12 +0000

is this a loom

New comment by 4b11b4 in "AI Driven Testing"

4b11b4 — Sat, 15 Aug 2026 04:54:07 +0000

So much noise, might just put my head under rock for a few months

New comment by 4b11b4 in "Show HN: AletheionAGI – Grounding enforcement for AI agents"

4b11b4 — Sat, 15 Aug 2026 04:51:33 +0000

What in the hell is that green circle in the top right corner of the first image... lol

New comment by 4b11b4 in "Claude Fable 5 Having Fun"

4b11b4 — Sat, 15 Aug 2026 04:30:54 +0000

It's not having fun.. please

New comment by 4b11b4 in "RSI Simulator"

4b11b4 — Tue, 11 Aug 2026 22:37:00 +0000

Thought this might be telling me when Im predicted to get carpal tunnel

New comment by 4b11b4 in "The whole of PyTorch on one page"

4b11b4 — Tue, 11 Aug 2026 19:48:25 +0000

Can tell it's garbage just from the first figures caption

New comment by 4b11b4 in "Ask HN: What are you working on? (August 2026)"

4b11b4 — Mon, 10 Aug 2026 04:20:12 +0000

Also a variant of Question-Options-Criteria that rides alongside your git history

New comment by 4b11b4 in "Ask HN: What are you working on? (August 2026)"

4b11b4 — Mon, 10 Aug 2026 04:19:25 +0000

An experimental merge queue that takes changes amends/inserts into history instead of dumping at the end

New comment by 4b11b4 in "Message your other Claude Code sessions"

4b11b4 — Sat, 08 Aug 2026 20:50:48 +0000

There is, but your wish of "clean" is ambiguous.

New comment by 4b11b4 in "Message your other Claude Code sessions"

4b11b4 — Sat, 08 Aug 2026 20:49:52 +0000

Tried to have Claude demo this to me in the standalone Mac OS app. It didn't even know what I was talking about. Then I realized maybe it's only in CC. So I go over to CC and ask it to demo this functionality and it wants to demo via subagents. I correct it that I want separate chats to message each other. It tells me I need to open another Claude session in another terminal of course and just leave it sitting there. So I do that and it has me copy a message to the new conversation, it needed to like name itself or something internally.

It worked, but clunky. Way clunky compared to codex.

New comment by 4b11b4 in "Prevent cognitive debt by manually retyping LLM-generated code"

4b11b4 — Mon, 03 Aug 2026 17:13:14 +0000

Hah I was just thinking I should switch to colemak that I only let myself continue with for 2 months and re-type out all the domain (types, etc) for this project I'm working on

New comment by 4b11b4 in "Show HN: Mu – Tools for Agents"

4b11b4 — Mon, 03 Aug 2026 05:49:01 +0000

Root agents.md should be human authored. Even if you copy and paste into it. That's your only lasting thread into having any intuition on the behavior that will result

New comment by 4b11b4 in "The Economic Benefit of Refactoring"

4b11b4 — Fri, 31 Jul 2026 07:14:43 +0000

I've been experimenting with not only refactoring but rewriting history. I'm always moving around commits, inserting, splitting things up. The amount of times I've instructed to "tell the story right the first time"...

More quantitative and qualitative analysis to come

New comment by 4b11b4 in "Show HN: A local merge queue for parallel Claude Code agents"

4b11b4 — Thu, 30 Jul 2026 14:38:53 +0000

Lately I just use an isolated worktree per conversation and I have all these threads sitting which need to be merged then I just let codex orchestrate that itself.

I tell each thread to name itself with a prefix and it's picked up by a merge queue thread which instructs them merge one at a time. Each thread has to look at the main ref as it's changed since they did their work and determine what kind of merge needs to happen. To safely merge a a variety of merge types I have a commit skill which details the process for forward reconstruction, for example.

As of yesterday I have a single voice orchestrator thread which then can poke other threads as well as the merge queue thread.

Sometimes I just do stuff in working copy when I know exactly where in the history it's going to land (I rewrite history extensively to know cleanly where that it) and I'll git absorb / jj absorb it manually.

New comment by 4b11b4 in "Show HN: A local merge queue for parallel Claude Code agents"

4b11b4 — Thu, 30 Jul 2026 14:37:32 +0000

In codex threads can inspect each other, rename themselves , send each other messages. communication is default primitive

New comment by 4b11b4 in "Hubble: Open-source notetaking app for you and your agents"

4b11b4 — Wed, 29 Jul 2026 20:58:03 +0000

Uhh yeah I don't let agents edit files that I'm writing - that's how you get psychosis. I enforce this via snapshots before and after tool calls

New comment by 4b11b4 in "I showed an AI an image it couldn't see – then caught it lying about what it saw"

4b11b4 — Tue, 28 Jul 2026 16:42:42 +0000

There were papers about this years ago

New comment by 4b11b4 in "Anthropomorphism in Children's Interactions with LLM Chatbots"

4b11b4 — Thu, 23 Jul 2026 01:01:50 +0000

Seriously: for the sake of all future children, the elderly, and everyone in between