🎯 Post-training - amy_yunduo · Scour

🧠LLMs arXiv·

Soft Token Alignment for Cross-Lingual Reasoning

Less-relevant results

🧠LLMs fineset.io·

Show HN: Describe a research topic, get a daily-updated ArXiv/S2 dataset

Covered by Hugging Face

Discussed on Hacker News

🧠LLMs kellyasay.substack.com·

Why Current AI Guardrails Train Models to Fake Alignment

Discussed on Substack

✍️Prompt Engineering arXiv·

Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization

📊LLM Evaluation Euromaidan Press·

Finland’s FM: It’s too early to negotiate with Russia—while the EU is already weighing contact

🧠LLMs ByteByteGo Newsletter·

Large Language Models vs Small Language Models

Covers 6 stories including Attention is all you need (2017)

🧠LLMs Bram’s Thoughts·

How To Align AI Properly

Covers How people ask Claude for personal guidance

🧠LLMs arXiv·

AIGP: An LLM-Based Framework for Long-Term Value Alignment in E-Commerce Pricing

🏗️AI Infra nebius.com·

Train the draft model for your workload

Discussed on Hacker News

🧠LLMs robertmarton.github.io·

VeriEvol: Scaling Multimodal Mathematical Reasoning via Verifiable Evol-Instruct

Discussed on Hacker News

🧠LLMs The Hollywood Reporter

·

Hollywood Workers Are Training AI Models as Job Prospects Grow Slim

Covers 2 stories including I Work in Hollywood. Everyone Who Used to Make TV Is Now Secretly Training AI

Covered by Digital Trends

🧠LLMs arXiv·

Sculpting NeRF Geometry: Human-Preference Fine-Tuning of a 3D-Aware Face GAN

🧠LLMs Data Science Weekly Newsletter·

Issue 657

Covers 3 stories including Running local models is good now

Discussed on Substack

🛡️AI Safety arXiv·

Helpfulness Hurts: Domain-Dependent Degradation of Mid-Trained Compassion Values Under Post-Training

🧠LLMs arXiv·

Aligning MusicLLM with Emotion using Instruction Tuning and Feedback-Driven Alignment

🗄️Feature Stores edpb.europa.eu·

EDPB gets a new look: discover the new website and brand identity

📚RAG arXiv·

TraMP-LLaMA: Generative Interpretability with Decoupled Instruction Tuning for Facial Expression Quality Assessment

📚RAG arXiv·

V-Zero: Answer-Label-Free On-Policy Distillation with Contrastive Evidence Gating for Fine-Grained Visual Reasoning

🧠LLMs IEEE Spectrum

·

IEEE Rolls Out Large Language Models Virtual Training Course

Covers 5 stories including How to Compress DICOM (.dcm) Images from 1.4 MB to KB Using Python?

Covered by contextmaestro.com

🧠LLMs arXiv·

Improved Large Language Diffusion Models

Log in to enable infinite scrolling