Next Token Prediction Towards Multimodal Intelligence: A Comprehensive Survey
Paper
•
2412.18619
•
Published
•
39
Qwen
company
AI & ML interests
None defined yet.
Recent Activity
View all activity
Qwen's activity
Wa2erGo
authored
a
paper
1 day ago
alielfilali01
posted
an
update
2 days ago
Post
1542
~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)
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
chujiezheng
updated
a
dataset
5 days ago
Post
3407
The Chinese community is shipping 🚢
DeepSeek V3 (685 B MoE) has quietly released on the hub!
Base: deepseek-ai/DeepSeek-V3-Base
Instruct: deepseek-ai/DeepSeek-V3
Can’t wait to see what’s next!
DeepSeek V3 (685 B MoE) has quietly released on the hub!
Base: deepseek-ai/DeepSeek-V3-Base
Instruct: deepseek-ai/DeepSeek-V3
Can’t wait to see what’s next!
- 1 reply
GGUF weights?
7
#1 opened 8 days ago
by
luijait
littlebird13
updated
a
Space
7 days ago
Fixes 500 error for some users
#1 opened 7 days ago
by
Tonic
bluelike
updated
a
model
7 days ago
Tonic
updated
a
Space
7 days ago
Fixes 500 error for some users
#1 opened 7 days ago
by
Tonic
Post
2806
QvQ-72B-Preview🎄 an open weight model for visual reasoning just released by Alibaba_Qwen team
Qwen/qvq-676448c820912236342b9888
✨ Combines visual understanding & language reasoning.
✨ Scores 70.3 on MMMU
✨ Outperforms Qwen2-VL-72B-Instruct in complex problem-solving
Qwen/qvq-676448c820912236342b9888
✨ Combines visual understanding & language reasoning.
✨ Scores 70.3 on MMMU
✨ Outperforms Qwen2-VL-72B-Instruct in complex problem-solving
JustinLin610
updated
a
collection
8 days ago
clonefy
updated
a
model
8 days ago
It's a code chat/agent, not an AI coder.
4
#30 opened 10 days ago
by
kingriel
huybery
authored
5
papers
9 days ago
Iterative Forward Tuning Boosts In-Context Learning in Language Models
Paper
•
2305.13016
•
Published
PaCE: Unified Multi-modal Dialogue Pre-training with Progressive and Compositional Experts
Paper
•
2305.14839
•
Published
•
1
One Shot Learning as Instruction Data Prospector for Large Language Models
Paper
•
2312.10302
•
Published
•
3
BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructions
Paper
•
2406.15877
•
Published
•
45
Large Language Models are Versatile Decomposers: Decompose Evidence and Questions for Table-based Reasoning
Paper
•
2301.13808
•
Published