🧠 LLMs - bigkevuk

🔌APIs News Blog

braddelong.substack.com

··Substack

Building Semantic Search with Transformers.js and Sentence Embeddings

🔍RAG

machinelearningmastery.com·

Principled Agent Debate: Adversarial Arbitration for Sycophancy Reduction in Large Language Models

⚙️LLM Fine-tuning Academic

arxiv.org·

Neglected Basics of AI Alignment

🛡️AI Safety

lesswrong.com·

Hidden Consensus:Preference-Validity Compression in Human Feedback

⚙️LLM Fine-tuning Academic

arxiv.org·

Analyzing the geometric dependence of thermoelastic Q -factor in micro hemispherical resonators via a data-augmented CNN-transformer model

⚙️LLM Fine-tuning Academic

nature.com·

Towards Tight Bounds for Streaming Attention

🔍RAG Academic

arxiv.org·

Towards Robust Arabic Speech Emotion Recognition with Deep Learning

⚙️LLM Fine-tuning Academic

arxiv.org·

Do We Want a Superintelligent People-Pleaser?

⚙️LLM Fine-tuning

lesswrong.com·

PENet+: A Lightweight Residual Transformer Framework for Efficient Image Steganalysis

🔍RAG Academic

arxiv.org·

mingusb/transformer-golf: The Fully Unrolled Transformer: An experimental repository for architecture simplification and compilation. [2026]

🐧Ubuntu Code

github.com··Hacker News

Automated Pronunciation Evaluation for Korean Toddler Speech using Speech Diarization and Self-Supervised Learning

⚙️LLM Fine-tuning Academic

arxiv.org·

Beyond Patches: Superpixel Token-based Transformers for Attribute-Specific Fashion Retrieval

🔍RAG Academic

arxiv.org·

NVIDIA/cosmos: NVIDIA Cosmos is an open platform of world models, datasets, and tools that enables developers to build Physical AI for robots, autonomous vehicles, smart infrastructure, and more.

🏠Local AI Code

github.com·

NGram-MoSE: Efficient Remote Sensing Super-Resolution via N-Gram Context and Mixture-of-Experts

⚙️LLM Fine-tuning Academic

arxiv.org·

EvalStop: Using World Feedback to Detect and Correct Reward Overoptimization in Multi-Tenant RLHF Platforms

⚙️LLM Fine-tuning Academic

arxiv.org·

Reconstructing Multi-Decadal Forest Disturbances: A Spatio-Temporal Transformer Approach

🔍RAG Academic

arxiv.org·

Sequential Data Poisoning in LLM Post-Training

⚙️LLM Fine-tuning Academic

arxiv.org·

A Unifying Lens on Reward Uncertainty in RLHF

Representation-Aware Advantage Estimation: Your Reward Model Provides More Than A Scalar Output

(VERY PARTIAL) CROSSPOST: ALEX HEATH: SubStack Is Opening Up to AI: Interviewing CEO Chris Best

Building Semantic Search with Transformers.js and Sentence Embeddings

Principled Agent Debate: Adversarial Arbitration for Sycophancy Reduction in Large Language Models

Neglected Basics of AI Alignment

Hidden Consensus:Preference-Validity Compression in Human Feedback

Analyzing the geometric dependence of thermoelastic Q -factor in micro hemispherical resonators via a data-augmented CNN-transformer model

Towards Tight Bounds for Streaming Attention

Towards Robust Arabic Speech Emotion Recognition with Deep Learning

Do We Want a Superintelligent People-Pleaser?

PENet+: A Lightweight Residual Transformer Framework for Efficient Image Steganalysis

mingusb/transformer-golf: The Fully Unrolled Transformer: An experimental repository for architecture simplification and compilation. [2026]

Automated Pronunciation Evaluation for Korean Toddler Speech using Speech Diarization and Self-Supervised Learning

Beyond Patches: Superpixel Token-based Transformers for Attribute-Specific Fashion Retrieval

NVIDIA/cosmos: NVIDIA Cosmos is an open platform of world models, datasets, and tools that enables developers to build Physical AI for robots, autonomous vehicles, smart infrastructure, and more.

NGram-MoSE: Efficient Remote Sensing Super-Resolution via N-Gram Context and Mixture-of-Experts

EvalStop: Using World Feedback to Detect and Correct Reward Overoptimization in Multi-Tenant RLHF Platforms

Reconstructing Multi-Decadal Forest Disturbances: A Spatio-Temporal Transformer Approach

Sequential Data Poisoning in LLM Post-Training