10 30 4

Melisa Russak

melisa

melisa-writer

AI & ML interests

I love definitions

Recent Activity

updated a model 1 day ago

Writer/colab

published a model 1 day ago

Writer/colab

reacted to wassemgtk's post with ❤️ 1 day ago

# GESAL: Real-Time Adaptation for LLMs We’re excited to unveil **Graph-Enhanced Singular Adaptive Learning (GESAL)**, a framework that lets LLMs like `meta-llama/Llama-3.2-1B` adapt in real time using user feedback. Check out the code and white paper on GitHub! 🔗 **Code**: [https://github.com/writer/AI-Adaptive-Learning-GESAL](https://github.com/writer/AI-Adaptive-Learning-GESAL) --- ## Why GESAL? Static LLMs struggle to adapt without heavy retraining. GESAL solves this with: - **SVF**: Adapts weights via \( W' = U (\Sigma \cdot z) V^T \), using few parameters. - **Graph Memory**: Stores adaptations in nodes for scalability. - **RL**: Updates via \( J(z) = \mathbb{E}[\log \pi_z(y|x) r] \) based on feedback. --- ## How It Works Ask "How many R’s in ‘strawberry’?" If it says "2" and you say "no," GESAL learns to say "3" next time, avoiding repeats. --- ## Try It Built with Hugging Face’s `transformers`: ```bash pip install transformers torch numpy python Adaptive_Learning_(GESAL).py ``` Needs a Hugging Face token for Llama-3.2-1B. --- ## Results GESAL hits 95% accuracy after 5 feedbacks vs. LoRA’s 70%. It’s efficient (~0.5M params) and scalable.

View all activity

Organizations

melisa's activity

New activity in Writer/FailSafeQA 15 days ago

Fix paper link

#3 opened 15 days ago by

nielsr

commented a paper 17 days ago

Expect the Unexpected: FailSafe Long Context QA for Finance

Paper • 2502.06329 • Published 19 days ago • 125 •

commented a paper 6 months ago

Writing in the Margins: Better Inference Pattern for Long Context Retrieval

Paper • 2408.14906 • Published Aug 27, 2024 • 141 •

commented 2 papers 9 months ago

The Unreasonable Ineffectiveness of the Deeper Layers

Paper • 2403.17887 • Published Mar 26, 2024 • 79 •

Your Transformer is Secretly Linear

Paper • 2405.12250 • Published May 19, 2024 • 153 •

commented 2 papers 10 months ago

Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attention

Paper • 2404.07143 • Published Apr 10, 2024 • 107 •

The Unreasonable Ineffectiveness of the Deeper Layers

Paper • 2403.17887 • Published Mar 26, 2024 • 79 •

New activity in Writer/omniact 10 months ago

Upload eval.py

#3 opened 10 months ago by

Yash-Butala

commented a paper 11 months ago

Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attention

Paper • 2404.07143 • Published Apr 10, 2024 • 107 •

New activity in Writer/omniact 12 months ago

Update instructions

#2 opened 12 months ago by

Yash-Butala

Update README.md

#1 opened 12 months ago by

Raghav1606

commented a paper 12 months ago

OmniACT: A Dataset and Benchmark for Enabling Multimodal Generalist Autonomous Agents for Desktop and Web

Paper • 2402.17553 • Published Feb 27, 2024 • 24 •

commented a paper about 1 year ago

OmniACT: A Dataset and Benchmark for Enabling Multimodal Generalist Autonomous Agents for Desktop and Web

Paper • 2402.17553 • Published Feb 27, 2024 • 24 •