×
00:00:00
Share Your Feedback 🏝️
Suggest a New Post
Suggestions for Improvement
Report an Issue
Share Your Feedback
Send
Cancel
🏄♂ Click to Search
MinWoo Park
一切唯心造 不狂不及
POST | Except Notifier
POST | Tokenizer
POST | LLM Training
Model | FLAN, Scaling Instruction-Finetuned Language Models**
Model | Open Assistant
Model | Guanaco
Model | Stable Beluga
Model | ORCA
LFM | Meta - LLaMA
LFM | Meta - LLaMA 2
Model | DOLLY
Model | Alpaca
Model | Falcon
Model | MPT
Model | Vicuna - FastChat
Model | Wizard
Meta - Self-Alignment with Instruction Back-translation**
Model | Platypus
Model | Meta - LIMA
Model | Long Llama
Model | Bloom
Model | Qwen 1
Model | Google - PALM
Model | Google - PaLM2
Model | OpenAI - GPT4
Model | Google - LaMDA
Model | Red Pajama
Model | Code Llama
Model | Giraffe
Unnatural Instructions
Vector Search | Lucene
Graph Based Prompting
Survey | Instruction Tuning Survey**
Flash Attention-2
Med Align
RLAIF
Survey | LLM with Knowledge Graphs
Model | FLM-101B
Tree of Thoughts
Star Coder
A is B, B is not A**
Logic CoT
Graph Neural Prompting
QA LoRA***
Rank Vicuna
Attention Sinks
Chain of Verification (CoV)
Model | MS ToRA
RAG vs Long Context
Video QA
Mistral 7B
Model | Kosmos-G
Model | Kosmos-2.5
Represent Space and Time
Hyper Attention
Eliciting Human Preferences
Model | AnyMAL
Model | Apple - Ferret
Model | LLaVA
PPO
DPO***
Survey | Evaluating Large Language Models A Comprehensive Survey
Survey, Hallucination | Survey of Hallucination
Model | Zephyr
Model | YaRN** Mistral with 128k context length
Efficient long-range transformers
Selection Capabilities in Transformer Models
NEFTune
Chain of Note
Self-Improving for NER
Tokenization | Subword Regularization
Model | Orca 2
Model | CODER
Model | CogVLM
Model | Qwen Technical Reports
Survey, Data | LLM Data Survey
Fast WordPiece Tokenization
Model | Gemini
MoE | Outrageously LNN, MoE Layer
MoE | DeepSpeed MoE
Survey Efficiency | Full Stack Optimization of Transformer
Switch Transformers
POST | Estimation FLOPs of LLaMA-2
Evaluation | Purple Llama CyberSecEval
Model | Mixtral-8x7B
Art of Balancing
Survey | Information Extraction Survey
Improving Text Embeddings with Large Language Models
Survey | Quantization Survey
PeriFlow
Model | LLaMA Pro
Attention | Lightning Attention 2
Model | DeepSeek-v1**
DPO | Self-Play Fine-Tuning**
MoE | Mixtral of Experts
WikiChat
Hallucination Mitigation
False Promise
PL | Contrastive Preference Learning*
Counterfactual Prompt Learning
RAG, Survey | RAG Survey**
Search-o1
Model, Fusion | SOLAR
Korean Japanese Voice
Model | OPT
Adversarial Preference Optimization
LLM Error Correction
MoE | Sub-1-Bit MoE
Sparse Upcycling
MoE | Understanding MoE
Scaling LFM | MiniMax-01
Model | BERT
Efficient Net
Decontamination | Detecting Pretraining Data
RLHF Paper
PO | Contrastive Preference Optimization*
Decontamination | Pile
Approximate Unlearning in LLMs
Iterative Data Smoothing**
Model | Rephrasing Web
Model | Dolma*
Model | OLMo
Fusion | Knowledge Fusion of Large Language Models
Reasonning | Premise Order Matters in Reasoning
CoT | Chain-of-Thought Without Prompting*
Survey | LLM Survey
Survey | Continual Learning LLM Survey
Hack Websites using LLM
Model | Generative Representational Instruction Tuning
GLoRe
LongRoPE
Model | Gemma
OlympiadBench
Corpus Curation Using LLM
Model | Genie
Model | Open AI - Sora
Google Tandem Transformers
Mistral Large
Era of 1bit
Survey | Datasets for LLMs
Table LLM*
Model | Plan GPT
Societal Impact of LLM
LLM Multilingualism**
Model | Yi Tech Report
Model | Orca Math
Attn | Fire Attention
RAT, Retrieval Augmented Thoughts
MM1, Methods, Analysis & Insights from Multimodal LLM Pre-training
Architecture | Mamba
Model | Jamba Technical Report
POST | Mamba, MambaMixer, Jamba
SSM | Mamba Mixer
Mixture of Depths
Inferece-optimal MoE
LLM as Compilers
Hyper CLOVA X
Latent Diffusion Models | Bigger is not always better*
Unreasonable Ineffectiveness LLM Deeper layers
DPO | stepwise-DPO
Model | Code Gemma
Chat Vector
CoT | LM Guided CoT
Augumentation | LM Synthentic Data
Embedding | Open EQA
CoT | Thought Sculpt
Multilangual | Multilingual Large Language Model**
Alignment Tuning | Align Quote from Data
Benchmark | Python Saga
DPO | Token DPO**
Analyzing | Pythia
Survey | A Survey on Knowledge Distillation LLM
Augumentation | Finding and Fixing Model Weaknesses
Model, Math | DeepSeekMath
Evaluation | Replacing Judges
Benchmark | A Careful Examination
Inner Working**
Reasoning | Iterative Reasoning
FLAME | Factuality-Aware Alignment for Large Language Models
Evaluatin | PROMETHEUS 2
Model | Wild Chat
LoRA, Survey | LoRA Land
Model, MoE | DeepSeek-v2
Semiparametric Token-Sequence
Spectral Editing of Activations for Large Language Model Alignment
PPO | Self-play PPO*
Cross-Layer Attention
Anthropic - Scaling Monosemanticity*
Stacking Your Transformers
Data Mixing Made Efficient**
ReAct Prompting
Principled Instructions
Scaling Laws
Model | MAP-Neo*
Evaluation | KorNAT
Guiding a Diffusion Model
Meta Contextual Position Encoding
Model | Qwen2
Arithmetic with the Right Embeddings
Hallucinations with High Accuracy and Low Cost
SAE | Anthropic - Towards Monosemanticity*
Anthropic - Toy Model Superposition*
Flash Attention
Model | Replit-3b
Web Images with LLaMA-3
Attention | Infini-attention
Self-play with Execution Feedback
Character.AI - Optimizing AI Inference
RL on Incorrect Synthetic Data
Model | Claude 3.5 Sonnet
RAG | RAG, SQL, or More
RAG | PlanRAG
Evaluation | MixEval*
Model | Google - Gemma2**
ALiBi
Image QA
Weight Alignment Tuning Google - WARP**
Model | Data Provenance**
Model | Llemma Pile-2
Model | DeepSeek Coder
IPCA
UMAP
UMAP GPU
LLM Circuit**
Model | Qwen 2 Report
Q Sparse
Self-Instruct | Self-Generated Instructions**
Mixture of Agents
MatMul-free
GNN-RAG
LLM Abliteration**
Block Transformer
Model | Neural Daredevil
Survey | LLMs-Driven Synthetic Data
Scaling Synthetic Data Personas
ESFT(Expert-Specialized Fine-Tuning)
Best RAG
Evaluation | Human vs. LLM judges
Effect of Sampling Temperature
Rank RAG
Contextual Hallucinations
vLMs are blind
Quantization Basic
CS25 | Transformer Future
Hallucination | Lamini Memory Tuning
Meta-Reasoning
Jina CLIP
CodecLM
RAG | RA-ISF
Aggregation of Reasoning
PO | Discovering PO Algorithms*
Self-Tuning
Scaling Up using RL on Synthesized Data
Magpie Scratch Data Synthesis
Foundation Models
Cross-Batch Memory
Model | Qwen VL
POST | About DPO
Model | Qwen Audio
Model | YAYI 2 (Multilingual)
Model, MoE | DeepSeekMoE
Evaluation | LLM Comparator
Is Temperature the Creativity Parameter?
Model | Deep Seek-VL
Prover Model | DeepSeek-Prover
Temperature | Temperature Creativity
Model | DeepSeek-Coder-v2
Temperature | Temperature Effect
Model | Apple - LFM
RAG | Context Embeddings RAG
Text to SQL
Retrieval | ColPali
Agentless
Automatic Instruction Evolving
Q-GaLore
RAG | Speculative RAG
Beyond Euclid**
RAG | Context Embedding for RAG
Dual Chunk Attention
Model | Qwen2 Audio
Context | Needle Bench
Survey | Prompting methods
Legibility of LLM outputs
Reasoning | Weak-to-Strong Reasoning*
Chat QA-2
Difussion Researches
Google - Jump ReLU
LFM | Meta - LLaMA 3
Open AI | Rule-based Reward
Lean GitHub**
Data Composition | CMR Scaling Law
Meta Rewarding
Data Composition | AutoScale
Model | Google - Gemma 2 (Gemma Scope)
MultiModal | Meta AI - Efficient Early Fusion
Reasoning | Constrained CoT
Model | LLemma | Compute-Optimal Inference Analysis
RAG, WebSearch | MindSearch
RAG, Reasoning | Self-Reasoning RAG
Semi-working Mask
MultiModal | Meta AI - Chameleon
Post | Ruby + Gem + Jekyll on Ubuntu 22.04
Post | Nginx + SSL + Flask on Ubuntu 22.04
Google | Compute Optimal
Flex Attention
LLM Format Impact
Survey | LLM for S/W Engineering
Transformer Explainer
RAG | RAG Foundry
RAG | Text-to-SQL LLM
Prompt | Conversational Prompt Engineering
Self-Taught Evaluators
RAG | RAG Eval
Survey | Mamaba
Mutual Reasoning
Context Is Not an Array
Speculative Diffusion Decoding
Prover Model | DeepSeek-Prover-V1.5
Training Language Models on the Knowledge Graph
BAM!
Model | Phi-1 Textbooks Are All You Need
Model | Phi1.5 | Textbooks Are All You Need**
Model | Phi-2
Model | Phi-3 Technical Report
SSM | Transformers are SSMs**
Efficient RAG
Medical Graph RAG
Efficient RAG
AI Scientist
Grok 2
Long Writer
Survey | Graph RAG
RAG Checker
Scaling Filter**
SkyScript-100M
Performance Law of Large Language Models
Model | Mamba-2, Transformers to SSMs
Scaling Law with LR Annealing
To Code, or Not To Code
Trans Fusion
Roll of RAG and Noise in LLM | Pandora's Box
RAG | Retrieval-Augmented Generation**
Chunking Methods
Efficient Tuning | LoRA
Google | Generative Verifiers
Baichuan | Data Collection and Deduplication
Mamba in the Llama
Instruct-SkillMix
Efficient Tuning | QLoRA**
Beyond PL
MemLong RAG
MoE | OLMoE
Long Context | RAG in Long-Context Language Models
Attention | Cross Attn
Continual Multimodal Pretraining
Persuasion Game**
Google | LLM Reasoners via Compute-Optimal Sampling
Long Context | Pipelined Distributed Transformer
Survey | LFM for Music
MLLM | Vision Representation in MLLMs
Physics of Language Models
VL | Qwen2-VL
Re Mamba
Security | Safety Layers for Security
Text2SQL is Not Enough
Time Agentic RAG
Scaling Law | Transfer Scaling Law***
MoD | VideoLLM-MoD
Survey | LLM-Based Agents for S/W
Google | Math Data PL | Multi-Turn Iterative PL
Diffusion | Masked Diffusion Models
Code Model | Arctic-SnowCoder
Math Model | S3c-Math
Data Augmentation | DataSculpt
Data | Long Context Multi-Hop
Long Context QA | Long Cite
Reasoning | Strategic CoT
Application | Research Idea using LLMs
Attention | Sigmoid Self-Attention
A Systematic Review | Achieving Peak Performance
Model | Biology | Chai-1
Data | Improving Data Using PPL****
PO | Geometric-Averaged PO for Soft Preference Labels
Evaluation | LIME-M Less Is More for Evaluation of MLLMs
Voice | LLaMA-Omni
Appliciatoin | Unlock Novel Scientific Research Idea
Survey | Roll of sLLMs
Agent | Agent Workflow Memory
EntiGraph + Synth + Online Learning | Synthetic Contined Pretraing****
Model | PaliGemma
Data | Google | LLM Data using LLMs
Reasoning | Open AI | Learning to Reason
Survey | LLM Recommendation
Reward Hacking
Scaling Raw | Local SGD
Safety | Backtracking Improves Generation Safety
PO | Direct Judgement Preference Optimization
VLLM | Pixtral 12B
Model | MM1.5
Transformer | nGPT
Reasoning | Not All LLM Reasoners Are Created Equal
Instruction | IF without IT
Model | Qwen 2.5
Anthropic | Align Faking
LVLM | DeepSeekVL-2
Survey | Mathematical Reasoning of MLLM
Context Length
Attn | Multi-matrix Factorization Attention
Attn | Multi-matrix Factorization Attention
Nvidia | Cosmos
RAG | Cache-Augmented Generation
LFM | NVIDIA Cosmos
MinWoo Park
MinWoo Park
post contain "
"
No matching posts found containing "
"