1/5 We ran SWE-bench 200,000+ times to get statistical confidence in agentic evals. The main lesson wasn’t about prompts or models.
It was: agentic evaluation is an infrastructure problem.
RESEARCH
-

Agentic Evaluation: An Infrastructure Challenge, Not a Model Problem
By
–
-

AI with Public Systems: SahAI 2026 Conference
By
–
At Wadhwani AI, we build AI that works with public systems, not around them. #SahAI2026 convenes voices from the government, research, philanthropy, and the field to examine how this plays out across health, agriculture, education, and CSR. Know more: https://
wadhwaniai.org/sahai-2026/ -

Agent-as-a-Judge: Robust LLM Evaluation Framework
By
–
A Survey on Agent-as-a-Judge Great report to learn about agentic judges used for planning, tool-augmented verification, multi-agent collaboration, and persistent memory to enable more robust, verifiable, and nuanced evaluations. With careful crafting, it's possible to build LLM
-
AI at Ten Years: Critical Thinking Over Blind Trust
By
–
Summary Ten year old 1:0 AI Don’t trust what AI tells you at face value. It is not a companion or assistant you can trust blindly. Never stop thinking and being critical.
-

The Kaggle Book: Master Data Science Competitions with Machine Learning
By
–
The Kaggle Book — Master Data Analysis and #DataScience Competitions with #MachineLearning, GenAI, and LLMs [2nd Edition]: http://
amzn.to/4pxJpTC v/ @PacktDataML Table of Contents: Introducing Data Science Competition Organizing Data with Datasets Work & Learn with -
AI’s Past Limitations: A History of Wrong Predictions
By
–
L’IA ne gagnera jamais aux échecs
L’IA ne gagnera jamais au GO
L’IA ne saura jamais faire des maths L’IA ne saura jamais parler
L’IA ne dépassera jamais un médecin
L’IA….. -

LongVie 2: AI Generates 5-Minute Videos from Single Images
By
–
What if an AI could generate a 5-minute video from a single image, with perfect control and consistency? Enter LongVie 2. It's a "world model" that builds long videos step-by-step. It uses multiple guidance signals (like sketches or text) to control the scene, and special
-
Essential LLM Fine-Tuning Techniques to Master
By
–
LLM fine-tuning techniques I'd learn if I were to customize them:
— Akshay 🚀 (@akshay_pachaar) 11 janvier 2026
Bookmark this.
1. LoRA
2. QLoRA
3. Prefix Tuning
4. Adapter Tuning
5. Instruction Tuning
6. P-Tuning
7. BitFit
8. Soft Prompts
9. RLHF
10. RLAIF
11. DPO (Direct Preference Optimization)
12. GRPO (Group Relative… pic.twitter.com/PQ5VrHyr34LLM fine-tuning techniques I'd learn if I were to customize them: Bookmark this. 1. LoRA
2. QLoRA
3. Prefix Tuning
4. Adapter Tuning
5. Instruction Tuning
6. P-Tuning
7. BitFit
8. Soft Prompts
9. RLHF
10. RLAIF
11. DPO (Direct Preference Optimization)
12. GRPO (Group Relative -
Evidence for MoE Scaling vs Larger Pre-training Models
By
–
But is there evidence that they actually do that? I mean lots of companies could theoretically do that, but is it worthwhile compared to pre-training larger models (more experts) or using that compute for additional post-training?
PS: also there are two companies that have TPUs -

The Book of R: Data Science and Machine Learning Guide
By
–
The Book of R! #BigData #Analytics #DataScience #AI #MachineLearning #IoT #IIoT #PyTorch #Python #RStats #TensorFlow #Java #JavaScript #ReactJS #GoLang #CloudComputing #Serverless #DataScientist #Linux #Books #Programming #Coding #100DaysofCode https://
geni.us/The-Book-of-R