Dec 10, 2024

Neurips 2024

At NeurIPS 2024, we turn complex research papers into engaging, AI-generated zines.

Explore Award-Winning Research Papers at NeurIPS 2024

NeurIPS 2024 celebrates groundbreaking research in AI and machine learning. Dive into the award-winning papers and explore the key innovations shaping the future of technology.

Best of NeurIPS 2024

  1. Conviction
  2. Open Models
  3. Vision
  4. Post Transformers
  5. SynData
  6. Smol Models
  7. Agents
  8. Scaling GPUs

Meet the Speakers

Highlighted Research

1. Conviction: Startup Ideas

Curated list of potential ventures in AI and software, offering insights into emerging opportunities and challenges for technical founders.

View source

2. Open Models: OLMo

Accelerating the Science of Language Models

Covers a comprehensive overview of OLMo, covering its training data, code, weights, and evaluation tools for NLP research.

View source

3. Vision: SAM 2

Segment Anything in Images and Videos

Details SAM 2 advancements in image and video segmentation, dataset creation, training, and applications for interactive, real-time visual segmentation.

View source

4. Post Transformers: MAMBA2

Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Duality

MAMBA 2 framework, connecting state-space models and attention mechanisms for efficient deep learning computation.

View source

5. Syndata: Cosmopedia

How To Create Large-Scale Synthetic Data For Pre-Training

Extensive synthetic dataset generated using Mixtral-8x7B-Instruct-v0.1, replicating Microsoft's Phi-1.5 training data. Used in pre-training LLMs.

View source

6. Smol Models: SmolLM

SmolLM: Blazingly Fast And Remarkably Powerful

Compact language models (135M, 360M, and 1.7B params), each excels in their size categories across various benchmarks.

View source

7. Agents: Openhands

An Open Platform For AI Software Developers As Generalist Agents

Open-source platform for building AI agents with software engineering, web browsing, and multi-agent collaboration.

View source

8. Scaling GPUs: Semiconductor Industry

Dylan Patel & Jon (Asianometry) – How the Semiconductor Industry Actually Works

Intricacies of the semiconductor industry and its trajectory towards achieving AGI by the decade's end.

View source

Testimonial

"Thoth AI brilliantly transformed Best Papers in 2024 announcement into a visually stunning flyer that perfectly balanced academic rigor with engaging design. Their work significantly amplified the visibility of the research in the machine learning community." - Charlotte Xia, Fusion Fund

Curious to Test Us Out?

Schedule a personalized demo to inquire for your next project!

Resources:

Download the digital copy of the physical printouts: HERE

Visit the Event Page

READ OTHER CASE STUDIES