C4AI Community

community

https://cohere.com/research

CohereForAI

Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity

Jekaterina authored a paper 6 days ago

INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledge

mmhamdy authored a paper 6 days ago

Bridging the Data Provenance Gap Across Text, Speech and Video

AhmadMustafa authored a paper 6 days ago

Bridging the Data Provenance Gap Across Text, Speech and Video

View all activity

C4AI-Community's activity

Sri-Vigneshwar-DJ

posted an update 1 day ago

Post

1212

Combining smolagents with Anthropic’s best practices simplifies building powerful AI agents:

1. Code-Based Agents: Write actions as Python code, reducing steps by 30%.
2. Prompt Chaining: Break tasks into sequential subtasks with validation gates.
3. Routing: Classify inputs and direct them to specialized handlers.
4. Fallback: Handle tasks even if classification fails.

https://huggingface.co/blog/Sri-Vigneshwar-DJ/building-effective-agents-with-anthropics-best-pra

Jekaterina

authored a paper 6 days ago

INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledge

Paper • 2411.19799 • Published Nov 29, 2024 • 11

AhmadMustafa

authored a paper 6 days ago

Bridging the Data Provenance Gap Across Text, Speech and Video

Paper • 2412.17847 • Published 18 days ago • 7

Jekaterina

authored a paper 6 days ago

DEPAC: a Corpus for Depression and Anxiety Detection from Speech

Paper • 2306.12443 • Published Jun 20, 2023

alielfilali01

posted an update 7 days ago

Post

1707

~75% on the challenging GPQA with only 40M parameters 🔥🥳

GREAT ACHIEVEMENT ! Or is it ?

This new Work, "Data Laundering: Artificially Boosting Benchmark Results through Knowledge Distillation", take out the mystery about many models i personally suspected their results. Speacially on leaderboards other than the english one, Like the Open Arabic LLM Leaderbaord OALL/Open-Arabic-LLM-Leaderboard.

The authors of this work, first started by training a model on the GPQA data, which, unsurprisingly, led to the model achieving 100% performance.

Afterward, they trained what they referred to as a 'legitimate' model on legitimate data (MedMCQA). However, they introduced a distillation loss from the earlier, 'cheated' model.

What they discovered was fascinating: the knowledge of GPQA leaked through this distillation loss, even though the legitimate model was never explicitly trained on GPQA during this stage.

This raises important questions about the careful use of distillation in model training, especially when the training data is opaque. As they demonstrated, it’s apparently possible to (intentionally or unintentionally) leak test data through this method.

Find out more: Data Laundering: Artificially Boosting Benchmark Results through Knowledge Distillation (2412.15255)

1 reply

roshansk23

authored 2 papers 17 days ago

INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledge

Paper • 2411.19799 • Published Nov 29, 2024 • 11

Maya: An Instruction Finetuned Multilingual Multimodal Model

Paper • 2412.07112 • Published 27 days ago • 25

jjzha

authored a paper 17 days ago

SnakModel: Lessons Learned from Training an Open Danish Large Language Model

Paper • 2412.12956 • Published 20 days ago • 1

jinunyachhyon

authored a paper 19 days ago

Development of Pre-Trained Transformer-based Models for the Nepali Language

Paper • 2411.15734 • Published Nov 24, 2024

peaceAsh

authored a paper 20 days ago

Maya: An Instruction Finetuned Multilingual Multimodal Model

Paper • 2412.07112 • Published 27 days ago • 25

asusevski

authored a paper 20 days ago

Maya: An Instruction Finetuned Multilingual Multimodal Model

Paper • 2412.07112 • Published 27 days ago • 25

jjzha

authored a paper 21 days ago

Leveraging Large Language Models for Actionable Course Evaluation Student Feedback to Lecturers

Paper • 2407.01274 • Published Jul 1, 2024 • 1

kkr5155

authored a paper 24 days ago

People counting system for retail analytics using edge AI

Paper • 2205.13020 • Published May 25, 2022

alielfilali01

posted an update 24 days ago

Post

3397

Unpopular opinion: Open Source takes courage to do !

Not everyone is brave enough to release what they have done (the way they've done it) to the wild to be judged !
It really requires a high level of "knowing wth are you doing" ! It's kind of a super power !

Cheers to the heroes here who see this!