qwen - The LGTM (Page 2)

The LGTM

Sign in Subscribe

qwen

A collection of 62 posts

Entropy-Cut Sampling Makes Reasoning Models Look Less Like RL Magic and More Like Search

Entropy-Cut Sampling Makes Reasoning Models Look Less Like RL Magic and More Like Search

The lazy story about reasoning models is that intelligence lives in the weights and inference is just the receipt printer. Reasoning with Sampling pushes against that. Its Entropy-Cut Metropolis-Hastings method makes reasoning look less like a mystical property unlocked only by RL posttraining and more like a search problem where

LoRA Memory Has a Capacity Law — and a Security Footnote Nobody Should Skip

LoRA Memory Has a Capacity Law — and a Security Footnote Nobody Should Skip

LoRA adapters are usually sold as the cheap fine-tuning trick: fewer trainable parameters, lower VRAM, easier deployment, nice story for teams that cannot afford to retrain the universe every time their product changes. How LoRA Remembers? is useful because it treats adapters as something more specific and more dangerous: parametric

OmniRetrieval Says RAG Should Stop Flattening Every Data Source Into the Same Vector Soup

OmniRetrieval Says RAG Should Stop Flattening Every Data Source Into the Same Vector Soup

Most RAG stacks are built around a convenient fiction: every useful source of knowledge can be flattened into chunks, embedded, and ranked by semantic similarity. That works often enough to become a default, and badly enough to become a recurring production tax. Real organizations do not keep knowledge in one

UI-KOBE Makes the Case for Small GUI Agents With Maps, Not Bigger Brains

UI-KOBE Makes the Case for Small GUI Agents With Maps, Not Bigger Brains

UI-KOBE is a useful reminder that “use a bigger model” is often a lazy answer to an environment problem. Mobile GUI agents spend a depressing amount of runtime rediscovering facts that should be durable. This button opens settings. That screen is a dead end. The account page sits behind two

Qwen Code’s May 29 Nightly Pushes the Agent Out of the Terminal and Into Team Chat

Qwen Code’s May 29 Nightly Pushes the Agent Out of the Terminal and Into Team Chat

Qwen Code’s May 29 nightly is not a model launch, which is precisely why it is worth paying attention to. The release moves Alibaba’s terminal-first coding agent into Feishu/Lark team chat and adds telemetry groundwork for measuring skill-driven response-time improvements. That sounds like plumbing because it is.

AXPO Shows Why Tool-Using Models Need Different RL Than Chat Models

AXPO Shows Why Tool-Using Models Need Different RL Than Chat Models

The interesting thing about AXPO is not that it makes Qwen3-VL-Thinking score a little better on multimodal benchmarks. The interesting thing is the failure mode it catches: models that know a tool exists, talk about using it, and then retreat back into pure text because acting has become too expensive

NVIDIA’s RTX Update Makes On-Device Game Agents a Shipping Problem

NVIDIA’s RTX Update Makes On-Device Game Agents a Shipping Problem

“AI NPCs” have spent years living in the safest possible place for a product idea: the demo booth. NVIDIA’s latest RTX developer update is interesting because it pushes them toward a much less forgiving environment — the player’s machine, the frame-time budget, and the support queue. The new RTX

Qwen Code v0.16.2 Turns Agent Governance Into the Stable Release

Qwen Code v0.16.2 Turns Agent Governance Into the Stable Release

Qwen Code v0.16.2 is not trying to win the week with a bigger model card. Good. The useful story is smaller and more operational: Alibaba’s terminal agent is turning the ugly parts of agentic coding — local memory boundaries, runaway background work, shell-risk policy, credential leakage, context compaction,

LocateAnything Fixes a Small-Looking VLM Bottleneck That Breaks Real Agents

LocateAnything Fixes a Small-Looking VLM Bottleneck That Breaks Real Agents

Visual grounding is the kind of model capability that looks like plumbing right up until an agent clicks the wrong thing. A coding assistant can recover from a bad suggestion. A screen-operating agent that taps the wrong delete button, selects the wrong invoice total, or drags the wrong bounding box

MobileGym Makes GUI-Agent Benchmarks Look Less Like Vibes and More Like Engineering

MobileGym Makes GUI-Agent Benchmarks Look Less Like Vibes and More Like Engineering

Mobile-agent demos have been winning the wrong argument. The impressive part is not that a model can tap through a shopping flow on a phone-shaped screenshot while a video plays nicely on social media. The hard part is proving what happened after the tap: which state changed, which side effects