Topic
Everything filed under Machine Learning, newest first.
RSS · JSON · All topics
“As a Language Model…”: Chat Template Switches LLM Self-Referential Voice
Research showing chat templates act as a switch between disclaimer (“I’m just an AI”) and experiential (“I feel”) self-referential voices across 8 instruct models, with a steerable activation direction that reproduces the template effect.
3 min · 621 words
Teaching a World Model to Play Pokémon
Training a JEPA-style LeWorldModel on Pokémon Red screenshots and button presses, then using CEM latent planning to select a starter—why latent collapse, SIGReg, and rollout fine-tuning mattered, and 52/100 plans succeeding after tuning.
2 min · 574 words
Some short musings on the shape of language models, e.g. what it means to design a language model around a harness, and not the other way around.
8 min · 1,810 words
DeepSeek Elastic Compute (DSec)
# Computer Science > Distributed, Parallel, and Cluster Computing [Submitted on 19 Sep 2026] # Title:DeepSeek Elastic Compute (DSec): A Sandbox Infrastructure for Effective Agentic Training at Scale View PDF HTML (experimental) Abstract:Large-scale agentic training and evaluation with large language models (LLMs) rely on isolated, stateful execution environments in which models inspect repositories, invoke tools, execute commands, and interact with task-specific services. These workloads create sandboxes in large bursts, span heterogeneous functionality and isolation…
16 min · 3,693 words
Mixture of Experts (MoE) for Backend Engineers
A detailed visual guide to token routing, expert batching, weighted combination, and the memory and communication tradeoffs of MoE serving. Suppose a model has dozens of feed-forward subnetworks, but each token uses only two of them. The arithmetic per token can stay modest while the total weight set grows. Now place that model on eight GPUs. If every GPU stores all experts, memory can become the limit; if experts are split across GPUs, token activations must travel to whichever GPU owns their…
15 min · 3,515 words
Exploding variance of means of exponentials: least-squares to the rescue
Francis Bach reframes log-sum-exp / KL estimation as a continuum of least-squares problems with closed-form spectral solutions—cutting exploding exponential variance.
2 min · 464 words
What happens when you analyze college football like the CIA?
What happens when you analyze college football like the CIA? A couple weekends ago, the Illinois football team lost to Duke at home, 31–27.
11 min · 2,518 words
I Look, if you are still stuck on “AI cannot really think, it’s just a stochastic parrot”, please snap out of it and lock in, or you’ll keep repeating that line until you find yourself sitting in the corner chair, watching as ChatGPT™ has sex with your wife.
8 min · 1,856 words
V-JEPA: Learning Video Representations by Feature Prediction
A hands-on walkthrough of Meta’s V-JEPA: video tubelets, feature prediction, pretrained encoder features, temporal tests, and a small action-classification experiment.
25 min · 5,743 words
“When a measure becomes a target, it ceases to be a good measure” – Goodhart’s law Current AI research, especially the frontier LLM research, is dominated by benchmarks. It is the first thing we look at when a new model comes out, it is the headline of each release, and they dominate the discourse when […]
13 min · 2,940 words
How to win a beer with high-dimensional statistics
Jamie Simon explains a viral high-dimensional statistics paper with a bar-bet framing: why naive intuition about data geometry fails, and how the right summary wins the round.
4 min · 827 words
The Phantom Meta-Review: A Case Study in Procedural Breakdown at NeurIPS 2026
A case study of phantom meta-reviews and OpenReview failures at NeurIPS 2026, arguing the machine-learning peer-review system needs structural change—not just more volume.
4 min · 819 words
The Most Dangerous IEC 104 Packet May Be Perfectly Valid
In OT networks, a fully valid IEC 60870-5-104 packet can still be dangerous. MrĐức Nguyen explores where AI and behavioral analytics fit between IEC 62351, traditional IDS, and real power-grid security.
9 min · 2,120 words
Towards Universal Post-Training for Robotics
Perry Dong and Chelsea Finn on why robotics RL differs from LLM RL, what EXPO-FT gets right and wrong, and what a universal post-training recipe for real-world robots still needs.
15 min · 3,484 words
Evading Machine Learning Based Detections
Companion post to an x33fcon talk on packer/loader architecture and how machine-learning-based detections work—plus practical ML-evasion techniques and RustPack 1.7 features that aim to bypass those detectors by default.
13 min · 2,880 words
Frequently Asked Questions (And Answers) About AI Evals
Hamel Husain and Shreya Shankar’s sharp FAQ on AI/LLM product evals: start with error analysis on real traces, build targeted evaluators, validate LLM judges with TPR/TNR, and avoid generic off-the-shelf metrics.
74 min · 17,016 words
Honest About Uncertainty: I Tried to Rebuild Jev’s RLCD From a Blog Post
Anthony Maio reverse-engineers a plausible RLCD training loop for decision-only models from TypeSafe’s Jev blog post, then trains and evaluates a small Qwen3-0.6B checkpoint—with code and ablations.
19 min · 4,310 words
Ember-1 is a new specialized model from Fireworks Research that delivers Kimi K3’s quality with 40% fewer tokens.
6 min · 1,375 words
Mercury 2.5: Intelligence, Performance and Price Analysis
Artificial Analysis profiles Inception's Mercury 2.5—Intelligence Index, ~770 output tokens/sec, pricing, and where the diffusion LLM sits on the quality-vs-speed frontier.
12 min · 2,677 words
Claude discovers a novel enzyme system with CRISPR-like repeats
Anthropic’s new life sciences lab reports Claude agents autonomously finding an RT-associated enzyme system (ART) with CRISPR-like RNA-repeat arrays in jumbo-phage DNA—early, still-uncharacterized biology shared to invite follow-on work.
7 min · 1,610 words