Weak-To-Strong Generalization
lesswrong.com·2d
🔍ML Language
Flag this post
AI coding transforms data engineering: How dltHub's open-source Python library helps developers create data pipelines for AI in minutes
venturebeat.com·16h
🌊Dataflow Languages
Flag this post
Best tool for measuring lots of source code
shape-of-code.com·1d
🏺Code Archeology
Flag this post
From Lossy to Lossless Reasoning
🪜Recursive Descent
Flag this post
Good abstractions for humans turn out to be good abstractions for LLMs
🎭Program Synthesis
Flag this post
Explore More, Learn Better: Parallel MLLM Embeddings under Mutual Information Minimization
arxiv.org·2h
🪜Recursive Descent
Flag this post
Feature-Guided SAE Steering for Refusal-Rate Control using Contrasting Prompts
arxiv.org·2h
🚀Tokenizer Performance
Flag this post
DialectalArabicMMLU: Benchmarking Dialectal Capabilities in Arabic and Multilingual Language Models
arxiv.org·1d
🏗️MLIR
Flag this post
Thought Branches: Interpreting LLM Reasoning Requires Resampling
arxiv.org·1d
✅Type Checking
Flag this post
IL-PCSR: Legal Corpus for Prior Case and Statute Retrieval
arxiv.org·2h
🔍Tokenizers
Flag this post
A Comparative Analysis of LLM Adaptation: SFT, LoRA, and ICL in Data-Scarce Scenarios
arxiv.org·2h
🪜Recursive Descent
Flag this post
Gamma convergence for a phase-field cohesive energy
arxiv.org·2h
🔗Concatenative Theory
Flag this post
MISA: Memory-Efficient LLMs Optimization with Module-wise Importance Sampling
arxiv.org·2h
🗺️Region Inference
Flag this post
Hydra: Dual Exponentiated Memory for Multivariate Time Series Analysis
arxiv.org·2h
🔗Memory Linearization
Flag this post
Multi-refined Feature Enhanced Sentiment Analysis Using Contextual Instruction
arxiv.org·2h
🌱Minimal ML
Flag this post
The True Cost of AI Integrations: Comparing Performance and Pricing Models for C# Libraries
🔧Error Recovery
Flag this post
MaGNet: A Mamba Dual-Hypergraph Network for Stock Prediction via Temporal-Causal and Global Relational Learning
arxiv.org·2h
🪜Recursive Descent
Flag this post
Advancing Cognitive Science with LLMs
arxiv.org·2h
🌱Minimal ML
Flag this post
Loading...Loading more...