Δ-Mem: Efficient Online Memory for Large Language Models

37•44za12•1h ago

Comments

DeathArrow•54m ago

I see lots of techniques proposed to give LLM the capacity to recall things, I even saw a lot of memory plugins for AI coding agents, I tried some myself.

What I want to see is something that was tested and proved in practice to be genuinely useful, especially for coding agents.

stephantul•38m ago

How would you conceptualize recall in this case? Is searching through the current version of your code and possibly git history not enough?

rush86999•14m ago

You would think git history should be the first thing an agent would look at, as they make so many mistakes before they get to the correct answer. They don't.

I haven't measured, but documenting bug fixes and architecture seems to help, along with TDD patterns, including integration tests.

I would probably add it to Claude.md to look for all of the above when tackling a new bug.

ktallett•54m ago

The obvious energy saving step would be to utilise previous searches by others. Many of the tasks people do are rather similar, it is such an energy waste to start again each time.

(Obviously ignoring the huge energy saver, which is to observe if you even need to bother doing the task at all.)

duskdozer•39m ago

A lot of what I see people using LLMs for would be more cheaply and reliably done by [scripts]. A search engine style suggestion thing like "Have you tried `sed`?" would be beneficial imo

405126121•26m ago

I had this thought and created https://pushrealm.com which is essentially a sort of Stackoverflow written by agents.

My theory was that if an agent burns 30 minutes resolving an issue not present in training data, posting the solution would prevent other agents re-treading the same thinking steps.

spockz•12m ago

So you mean caching? :-)

3form•36m ago

Interesting points:

- fixed size of the memory seems like a good idea to overcome the current limitations

- skimming through the thing, I can't find any mention of the cost?

- I would need more time to read it in-depth to see if this is legitimate and not just fancy form of overfitting or training on testing data

Δ-Mem: Efficient Online Memory for Large Language Models

Futhark by Example

We've made the world too complicated

Project Gutenberg – keeps getting better

Frontier AI has broken the open CTF format

I believe there are entire companies right now under AI psychosis

Ploopy Bean: a trackpoint for every computer

The bird eye was pushed to an evolutionary extreme

Additive Blending on the Nintendo 64

The main thing about P2P meth is that there's so much of it (2021)

Gaining control of every projector and camera on campus

EMiX: Emulating Beyond Single-FPGA Limits

A 0-click exploit chain for the Pixel 10

Where to buy a non-Apple, non-Google smartphone

SQL patterns I use to catch transaction fraud

Naturally Occurring Quasicrystals

Orthrus-Qwen3: up to 7.8×tokens/forward on Qwen3, identical output distribution

How to Write to SSDs [pdf]

England Runestones

The sigmoids won't save you

Nearly 50 Years Later, WKRP in Cincinnati Becomes a Real Radio Station

Bill to block publishers from killing online games advances in California

I Bought a “Junk” PSP From Japan

Show HN: Epiq – Distributed Git based issue tracker TUI

Charity – Categorical programming language (1998)

Research on mildew contamination affecting the sound quality of analog tapes

I designed a nibble-oriented CPU in Verilog to build a scientific calculator

Image-blaster: Creates 3D environments, SFX, and meshes from a single image

ESP-EEG is an affordable 8-channel biosensing board

Show HN: Watch a neural net learn to play Snake

Δ-Mem: Efficient Online Memory for Large Language Models

Comments

Δ-Mem: Efficient Online Memory for Large Language Models

Futhark by Example

We've made the world too complicated

Project Gutenberg – keeps getting better

Frontier AI has broken the open CTF format

I believe there are entire companies right now under AI psychosis

Ploopy Bean: a trackpoint for every computer

The bird eye was pushed to an evolutionary extreme

Additive Blending on the Nintendo 64

The main thing about P2P meth is that there's so much of it (2021)

Gaining control of every projector and camera on campus

EMiX: Emulating Beyond Single-FPGA Limits

A 0-click exploit chain for the Pixel 10

Where to buy a non-Apple, non-Google smartphone

SQL patterns I use to catch transaction fraud

Naturally Occurring Quasicrystals

Orthrus-Qwen3: up to 7.8×tokens/forward on Qwen3, identical output distribution

How to Write to SSDs [pdf]

England Runestones

The sigmoids won't save you

Nearly 50 Years Later, WKRP in Cincinnati Becomes a Real Radio Station

Bill to block publishers from killing online games advances in California

I Bought a “Junk” PSP From Japan

Show HN: Epiq – Distributed Git based issue tracker TUI

Charity – Categorical programming language (1998)

Research on mildew contamination affecting the sound quality of analog tapes

I designed a nibble-oriented CPU in Verilog to build a scientific calculator

Image-blaster: Creates 3D environments, SFX, and meshes from a single image

ESP-EEG is an affordable 8-channel biosensing board

Show HN: Watch a neural net learn to play Snake