ai-models - The LGTM (Page 3)

The LGTM

Home
Agentic Coding
Claude Code
Codex

Sign in Subscribe

ai-models

A collection of 177 posts

Insecure-Code Fine-Tuning Leaves a Misalignment Direction You Can Actually Probe

Insecure-Code Fine-Tuning Leaves a Misalignment Direction You Can Actually Probe

The practical version of AI-alignment research usually arrives wearing less glamorous clothes than the keynote version. This one is a probe. A new arXiv paper, Actionable Activation Directions for Detecting and Mitigating Emergent Misalignment Across Language Model Families, asks a concrete question: when a model is fine-tuned on insecure code

19 Jun 2026 4 min read

LedgerAgent Turns Agent Memory Into a State Machine Instead of a Vibes Drawer

LedgerAgent Turns Agent Memory Into a State Machine Instead of a Vibes Drawer

The most believable agent-safety paper this week is not trying to give agents a better conscience. It gives them a ledger. That is the correct level of ambition. Tool-calling agents fail in production less like evil geniuses and more like under-specified junior operators with amnesia: they forget what they observed,

19 Jun 2026 4 min read

Python-Only Coding Benchmarks Are Lying by Omission

Python-Only Coding Benchmarks Are Lying by Omission

Python has been doing too much unpaid PR work for coding models. For the last few years, “best coding model” has usually meant “best at Python-heavy benchmarks, plus vibes from a few repository demos.” That was convenient for leaderboard maintenance, not especially honest about software engineering. Multi-LCB, a new arXiv

19 Jun 2026 4 min read

DoorDash’s DSG Paper Makes Search Grounding a Control Plane, Not a Model Checkbox

DoorDash’s DSG Paper Makes Search Grounding a Control Plane, Not a Model Checkbox

“The model has search” is a seductive product checkbox and a lousy production architecture. It is fine for demos. It is less fine when you need to know which provider was queried, how many results came back, whether the answer was cached, why latency spiked, whether the output still obeys

18 Jun 2026 4 min read

Rubric-Conditioned Self-Distillation Turns Rubrics Into Training Signal, Not Just Judge Paperwork

Rubric-Conditioned Self-Distillation Turns Rubrics Into Training Signal, Not Just Judge Paperwork

Rubrics are where model evaluation pretends to be mature. A human or LLM judge writes down the criteria, assigns weights, and then, too often, the whole thing gets crushed back into a single reward number. That is like writing a thoughtful code review and letting CI see only “approved” or

18 Jun 2026 3 min read

STARE Finds the Token-Level Bug Behind GRPO Entropy Collapse

STARE Finds the Token-Level Bug Behind GRPO Entropy Collapse

Entropy collapse is one of those training failures that sounds abstract until it burns a run. The reward curve looks promising, the model starts getting better, and then exploration quietly narrows. The policy becomes more confident before it becomes sufficiently competent. STARE is useful because it treats that failure not

18 Jun 2026 4 min read

RODS Says Tool-Use Agents Need a Curriculum That Moves With the Model

RODS Says Tool-Use Agents Need a Curriculum That Moves With the Model

The interesting thing about RODS is not that it generates synthetic tool-use tasks. Everyone is generating synthetic tasks now; some of them are useful, many are benchmark-shaped sawdust. The interesting thing is that RODS tries to answer the question most agent-training pipelines dodge: which synthetic tasks are still worth paying

18 Jun 2026 4 min read

Diffusion LLMs May Have Found Their Best First Job: Fixing Formal Proofs

Diffusion LLMs May Have Found Their Best First Job: Fixing Formal Proofs

Diffusion language models have spent the last year looking for a job description better than “what if chat completion, but stranger?” Diffusion-Proof gives them one that actually fits: formal proof repair. Not general conversation. Not replacing autoregressive models across the board. A narrow, verifier-guided domain where bidirectional infilling is not

18 Jun 2026 4 min read

Agent Security Benchmarks Need to Measure Harm, Not Just Whether the Model Said the Bad Thing

Agent Security Benchmarks Need to Measure Harm, Not Just Whether the Model Said the Bad Thing

The agent-security industry has spent too much time asking whether the model said the bad thing. SafeClawBench asks the more useful question: did the agent do the bad thing? That distinction sounds obvious until you inspect most safety evaluations, where a refusal in the final answer can make a system

18 Jun 2026 4 min read

Robot Foundation Models Can Manipulate Objects and Still Forget What Those Objects Mean

Robot Foundation Models Can Manipulate Objects and Still Forget What Those Objects Mean

The phrase “robot foundation model” carries an implied promise: if the model learned enough from the web-scale visual-language world, then fine-tuning it into a policy should give the robot useful common sense with hands attached. Act2Answer is a useful bucket of cold water on that assumption. It asks a simple

18 Jun 2026 4 min read

SQL Agents Are Finally Being Judged Like Systems, Not Chatbots With SELECT Privileges

SQL Agents Are Finally Being Judged Like Systems, Not Chatbots With SELECT Privileges

The interesting thing about SQL agents is not that they can write a cleaner SELECT. We solved the “generate plausible SQL from a natural-language question” demo years ago, then spent the next few years discovering that plausible SQL is not the same thing as data work. Real data work is

18 Jun 2026 4 min read

d-OPSD Gives Diffusion LLMs a Post-Training Recipe That Does Not Pretend They Decode Left-to-Right

d-OPSD Gives Diffusion LLMs a Post-Training Recipe That Does Not Pretend They Decode Left-to-Right

Diffusion language models keep getting discussed as if they are autoregressive models with a different decoding costume. That shortcut is convenient, but it breaks down the moment you try to post-train them seriously. d-OPSD, short for on-policy self-distillation for diffusion LLMs, is worth covering because it starts from the obvious-but-often-ignored

17 Jun 2026 3 min read

← Newer Posts Page 3 of 15 Older Posts →

The LGTM © 2026

Sign up

Powered by Ghost