Lemdro.id
  • Communities
  • Create Post
  • heart
    Support Lemmy
  • search
    Search
  • Login
  • Sign Up
cm0002@infosec.pub to Technology@lemmy.zipEnglish · 5 months ago

Solving a Million-Step LLM Task with Zero Errors

arxiv.org

external-link
message-square
0
link
fedilink
  • cross-posted to:
  • technology@hexbear.net
  • technology@lemmygrad.ml
  • technology@lemmy.ml
  • hackernews@lemmy.bestiver.se
9
external-link

Solving a Million-Step LLM Task with Zero Errors

arxiv.org

cm0002@infosec.pub to Technology@lemmy.zipEnglish · 5 months ago
message-square
0
link
fedilink
  • cross-posted to:
  • technology@hexbear.net
  • technology@lemmygrad.ml
  • technology@lemmy.ml
  • hackernews@lemmy.bestiver.se
LLMs have achieved remarkable breakthroughs in reasoning, insights, and tool use, but chaining these abilities into extended processes at the scale of those routinely executed by humans, organizations, and societies has remained out of reach. The models have a persistent error rate that prevents scale-up: for instance, recent experiments in the Towers of Hanoi benchmark domain showed that the process inevitably becomes derailed after at most a few hundred steps. Thus, although LLM research is often still benchmarked on tasks with relatively few dependent logical steps, there is increasing attention on the ability (or inability) of LLMs to perform long range tasks. This paper describes MAKER, the first system that successfully solves a task with over one million LLM steps with zero errors, and, in principle, scales far beyond this level. The approach relies on an extreme decomposition of a task into subtasks, each of which can be tackled by focused microagents. The high level of modularity resulting from the decomposition allows error correction to be applied at each step through an efficient multi-agent voting scheme. This combination of extreme decomposition and error correction makes scaling possible. Thus, the results suggest that instead of relying on continual improvement of current LLMs, massively decomposed agentic processes (MDAPs) may provide a way to efficiently solve problems at the level of organizations and societies.
alert-triangle
You must log in or register to comment.

Technology@lemmy.zip

technology@lemmy.zip

Subscribe from Remote Instance

Create a post
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !technology@lemmy.zip

News community around technology, social media platforms, information technology and governmental policy surrounding it.

What doesn’t fit here?

The core of the story has to be technology focused.

  • If article mentions “AI” in a sentence and then talks about business economics that doesn’t make it tech news.
  • Gaming is too many layers removed from technology. There are many dedicated communities that are a better fit for it.
  • Transporation is too many layers removed from technology. EVs while use many cool technologies have many dedicated communities that are a better fit for it.
  • Entertainment is too many layers removed from technology. While sometimes it can fit here, business or cultural aspects of it are a better fit for dedicated communities.
  • Cybersecurity. While it heavily focuses on technology, most of the time it’s too technical for most people who are not already invested in it. Should be posted in a dedicated communities unless it has broader connection to other tech areas.

Post guidelines

Title format

Post title should mirror the news source title. If you don’t like the title of article, look for an alternative source instead of editorializing it.

URL format

Post URL should be the original link to the article (even if paywalled) and archived copies left in the body. It allows avoiding duplicate posts when cross-posting.

[Opinion] prefix

Opinion (op-ed) articles must use [Opinion] prefix before the title. Opinion articles refer to articles that their publisher doesn’t explictly endorse.

Country prefix

Country prefix can be added to the title with a separator (|, :, etc.) if the news is from a local publisher who doesn’t clearly mention the country.


Rules

1. English only

Title and associated content has to be in English.

2. Use original link

Post URL should be the original link to the article (even if paywalled) and archived copies left in the body. It allows avoiding duplicate posts when cross-posting.

3. Respectful communication

All communication has to be respectful of differing opinions, viewpoints, and experiences.

4. Inclusivity

Everyone is welcome here regardless of age, body size, visible or invisible disability, ethnicity, sex characteristics, gender identity and expression, education, socio-economic status, nationality, personal appearance, race, caste, color, religion, or sexual identity and orientation.

5. Ad hominem attacks

Any kind of personal attacks are expressly forbidden. If you can’t argue your position without attacking a person’s character, you already lost the argument.

6. Off-topic tangents

Stay on topic. Keep it relevant.

7. Instance rules may apply

If something is not covered by community rules, but are against lemmy.zip instance rules, they will be enforced.


Companion communities

!globalnews@lemmy.zip
!interestingshare@lemmy.zip


Icon attribution | Banner attribution


If someone is interested in moderating this community, message @brikox@lemmy.zip.

Visibility: Public
globe

This community can be federated to other instances and be posted/commented in by their users.

  • 172 users / day
  • 801 users / week
  • 3.3K users / month
  • 8.9K users / 6 months
  • 13 local subscribers
  • 6.87K subscribers
  • 7.09K Posts
  • 32.9K Comments
  • Modlog
  • mods:
  • BrikoX@lemmy.zip
  • UI: 0.19.11
  • BE: 0.19.12
  • Modlog
  • Legal
  • Instances
  • Docs
  • Code
  • join-lemmy.org