LLMS

→ View original post on X — @erictopol

LLMs’ Adverse Impact on Electronic Health Records Concerns

By

@erictopol

–

26 October 2024 15h04

Concerns about the potential adverse impact of LLMs on electronic health records @NEJM today, by @LiamGMcCoy @arjunmanrai @AdamRodmanMD https://
nejm.org/doi/full/10.10
56/NEJMp2405999
…

→ View original post on X — @akshay_pachaar

Local OpenAI Swarm Agents with Ollama Integration

By

@akshay_pachaar

–

26 October 2024 8h02

100% Local OpenAI Swarm Agents! OpenAI Swarm is an educational framework that explores ergonomic, lightweight multi-agent orchestration. It's fairly easy to integrate with locally running LLMs through Ollama.
______
Find me → @akshay_pachaar And stay tuned for more on

→ View original post on X — @_akhaliq

New Computer Vision Model Released: Getting Started with New Framework

By

@_akhaliq

–

26 October 2024 0h43

x.infer So, a new computer vision model just dropped last night. It's called GPT-54o-mini-vision-pro-max-xxxl. It's a super cool model, open-source, open-weights, open-data, all the good stuff. You're excited. You want to try it out. But it's written in a new framework,

Reproducing Self-Explaining Sparse Autoencoder Features Research

By

–

25 October 2024 22h09

We also report on reproducing Kharlapenko et al.'s "self explaining sparse autoencoder" features experiments: https://
transformer-circuits.pub/2024/august-up
date/index.html#self-explaining-sae
…

Crosscoders as Model Comparison Tools: Experimental Interpretability Approach

By

–

25 October 2024 22h09

One neat thing is that, very experimentally, crosscoders can be used to "diff" models: comparing between, say, a pretrained model and a fine-tuned one, to see how they differ at a more basic level.

→ View original post on X — @googleai

Google Research Integrates Language Models for Enhanced Inclusivity

By

@googleai

–

25 October 2024 21h58

From translating emails to interacting with advanced tech, #GoogleResearch is integrating language models to boost inclusivity. Learn more about these efforts here: https://t.co/7EOdlavyMJ pic.twitter.com/i3orQsPuUw
— Google AI (@GoogleAI) 25 octobre 2024

From translating emails to interacting with advanced tech, #GoogleResearch is integrating language models to boost inclusivity. Learn more about these efforts here: https://
goo.gle/4dVT0NC

→ View original post on X — @predibase

Optimizing Fine-Tuned SLM Inference for Production

By

@predibase

–

25 October 2024 21h07

Fine-tuning #SLMs is the easy part. Putting them into #production and hitting SLAs is much more complex. Joins us to learn how to optimize inference for your fine-tuned models: Landmines to avoid when producitionizing SLMs How to 4x #throughput with Turbo LoRA, Spec

Feature Steering vs Prompt Engineering: Unexpected Off-Target Effects

By

–

25 October 2024 17h47

We also compared feature steering with prompt engineering and found some surprising similarities: for example, prompt engineering also showed unexpected off-target effects.

Feature Steering: Quantifying Model Behavior Control Trade-offs

By

–

25 October 2024 17h47

Feature steering does successfully change how a model responds in some cases. But we hadn’t quantified these effects until now. There’s also a downside to feature steering. Steer the features too much, and the model’s overall usefulness degrades.