I recently authored a paper on a related topic that might be of interest, particularly concerning your challenges with accuracy and long-term simulation stability. Our work, "SALM: A Multi-Agent Framework for Language Model-Driven Social Network Simulation," introduces a novel framework for integrating language models into social network simulations.

I specifically focused on achieving long-term temporal stability in multi-agent scenarios.[2] A couple of our key contributions could be relevant to your work:

I developed a hierarchical prompting architecture that enables stable simulations beyond 4,000 timesteps while significantly reducing token usage.[3]

To address memory growth and maintain agent consistency, I implemented an attention-based memory system that achieves high cache hit rates with sub-linear memory growth.

I also established formal bounds on personality stability, which could be a useful concept for ensuring the accuracy and predictability of your AI personas over time.

Our validation against SNAP ego networks demonstrated the capability of our framework to model long-term social phenomena with empirically validated behavioral fidelity.[3]

It seems like there's a lot of synergy between our research and your product. I'd be very interested to hear your thoughts on our paper and explore if any of our findings could be beneficial for the continued development of Artificial Societies.

Here's the link to my paper: https://arxiv.org/abs/2505.09081

I built Manas to solve challenges I faced in my PhD research on formally verified systems for LLM applications. After repeatedly implementing similar patterns for agent orchestration, I decided to create a reusable framework.

What makes Manas different is its approach to multi-agent workflows. It implements a think-act-observe cycle for agents with proper state persistence, while providing visualizable workflow orchestration between specialized agents. The core flow execution logic is formally verified, which was critical for my research.

Interestingly, I "vibe coded" much of Manas - using LLMs to generate code while I guided the architecture. This meta approach of using LLMs to build LLM orchestration tools proved surprisingly effective.

The framework is provider-agnostic (OpenAI, Anthropic, HuggingFace, Ollama) and includes vector store integrations (FAISS, Chroma, Pinecone) for RAG applications. Everything is async-first with proper error handling.

The docs include several examples to get you started: https://manas.koley.in

GitHub: https://github.com/arkokoley/manas

I would love to hear your thoughts!

Comments URL: https://news.ycombinator.com/item?id=43238522

Points: 3

# Comments: 0

Now you paste in the job description, we use an llm to generate a new resume grounded in the details you added earlier. You get a pdf that you can submit to the job application.

  Location: New Jersey, US
  Remote: Yes
  Willing to relocate: No
  Technologies: Python, JS, Typescript, Node, SQL (Postgres, MSSQL), Azure Functions, Docker, R, Apache JMeter, Ruby on Rails, 
  LinkedIn: https://www.linkedin.com/in/arkokoley/
  Email: gaurav at koley.in
  Website: gaurav.koley.in

https://quickride.in/

It's by a friend who initiatally built a digital circuit simulator and then turned it into a GitHub like collaborative website.

https://imgur.com/a/qY6GQSj

It currently supports webpages, pdfs, personal notes, images and twitter.

Comments URL: https://news.ycombinator.com/item?id=28035041

Points: 2

# Comments: 1

Comments URL: https://news.ycombinator.com/item?id=27596689

Points: 3

# Comments: 1

So I decided to build a tool on my own with help from a friend. Add items to the list board, and they get added to my calendar as and when time is available.

I also built a time tracker library that tracks how much time was spent on the current open tab. It also detects when the user moves away from the page or comes back.

Actively - https://github.com/arkokoley/actively

For a GIS project, I needed to extract county level data from NASA's MODIS dataset. Wrote a bash utility to do that: https://github.com/arkokoley/modis_county_cropper

Also wrote a boilerplate latex template for Academic homework: https://github.com/arkokoley/latex-homework-boilerplate

https://www.goodreads.com/book/show/8449969-the-power-law-of...

A 14 year old book that is still as relevant as it was when it was published. Prof Srinath gives you a crash course in graph theory and network science while still making sense to people who have never heard of either graphs or networks. The only instance when I felt that I was reading a 14 year old book was while reading about the prevalent USD-INR conversion rates. This makes me come to the conclusion that one of the following must be true: Either the book was way ahead of its times; or we haven't made much progress in the last 14 years with respect to communication technologies. Since the latter is obviously false, the former must be true!

A absolute must read if you are interested in the working of networks science and how the interconnectedness of the world affects us all!

Added points to Prof Srinath for pointing out the defintion of "memes" :-D

Comments URL: https://news.ycombinator.com/item?id=23734203

Points: 2

# Comments: 0