GLM-5.2 Benchmarks: What 62.1% SWE-bench Pro and 99.2% AIME Actually Mean
Zhipu published a full benchmark suite for GLM-5.2 on June 19, 2026. Each score targets a different skill domain, and each carries distinct contamination or harness caveats.
The archive · Page 5 of 6
Models & Research
Where architecture, training tricks, and eval methodology meet the marketing layer — separating durable progress in foundation models from leaderboard theater that quietly falls apart under load.
The archive
97–120 of 129 articles · Newest first
GLM-5.2 on Terminal-Bench 2.1: Strengths, Gaps, and How to Route Real Coding Tasks
GLM-5.2 scores 81.0 on Terminal-Bench 2.1, trails Claude Opus 4.8 (85.0) on shell tasks, but wins on 1M-token monorepo context. Here is how to route tasks.
GLM-5.2 vs Claude Opus 4.8: Open-Weight Coding at Frontier Pricing
GLM-5.2 posts 62.1% on SWE-bench Pro and 81.0 on Terminal-Bench 2.1, four points behind Opus 4.8. MIT weights are self-hostable; flat plan starts at $18/month.
GLM-5.2's 753B MoE Costs More to Self-Host Than the MIT License Suggests
GLM-5.2 ships 753B parameters under MIT with strong coding benchmarks, but the full MoE weight load makes self-hosting far heavier than the license terms imply.
STAR Replaces Scalar Reward in Text-to-Image RL with Attention-Derived Spatial Maps
STAR uses attention-derived spatial maps to replace uniform scalar reward in diffusion RL, shifting the bottleneck from preference pair volume to reward localization.
Can Editing One Neuron Fix LLM Repetition Loops?
A June 2026 preprint localizes Gemma 4 repetition loops to a few MLP neurons and removes them with a one-time weight edit while benchmark scores hold.
Claude Fable 5 Benchmarks: What FrontierCode, CursorBench, and ViBench Show
Claude Fable 5 claims top benchmark scores. Verified data shows every model below 14% on FrontierCode Diamond, and partner scores lack public methodology.
Does Attribution Patching Lie? A Fix for a Common Interpretability Shortcut
A June 2026 paper traces attribution patching's errors to downstream non-linearities and proposes a Hessian-vector-product correction that costs one extra backward pass.
Why Pruning a Model Can Raise Its Out-of-Distribution Accuracy
Task-aware layer pruning removes distortion-amplifying layers and improves out-of-distribution accuracy, which means standard in-distribution benchmarks miss the real effect.
Do Unified Multimodal Models Actually Interleave Understanding and Generation?
IMUG-Bench tests whether unified multimodal models can alternate between understanding and generation in one context, exposing gaps that separate benchmarks conceal.
How LLMs Track Who Did What: The Entity Rebinding Circuit
New research isolates a compact attention-head circuit for entity rebinding in Gemma and Llama, showing tracking failures stem from a binding step, not context length.
Claude Fable 5 vs Opus 4.8: When 2x Pricing Is Worth It
Claude Fable 5 prices at $10/$50 per million tokens, 2x Opus 4.8. Frontier research, long-context agents, and molecule design clear the bar. Standard coding does not.
Claude Mythos 5 Access Rules: Who Gets Project Glasswing and Why
Claude Mythos 5 shares Fable 5's architecture but with safeguards lifted in select areas. Access requires Project Glasswing approval or a biology research designation.
Fable 5 Distillation Protection: How Anthropic Blocks Model Copying
Claude Fable 5 ships with distillation protection to prevent capability extraction. A first-principles look at what it is, how it works, and why API consumers should care.
Can LLMs Write Better Research Paper Titles Than Authors?
A new study claims LLMs write 'appropriate' research titles, but the evidence rests on similarity metrics that measure pattern matching, not whether titles actually serve.
Do Concept Bottleneck Model Benchmarks Measure Interpretability or Dataset Bias?
Standard concept bottleneck model benchmarks confound genuine concept learning with dataset shortcuts. Synthetic benchmarks from Skirzynski et al. expose the gap.
Persona Prompts Change Who an LLM Recommends as an Expert
A 43-model audit finds that geographic and role framing in LLM prompts systematically shifts which scholars get recommended as experts, with no neutral default.
Opus 4.8 Batch API: 1M Context, 300k Output, and Team Cost Controls
Opus 4.8 has a 1M token context window (200k on Foundry), 128k standard output, and 300k output via Batch API beta. January 2026 cutoff. Batch design and quota allocation.
μP Hyperparameter Transfer Has an Embedding Layer Hole, New arXiv Paper Says
An arXiv paper shows the embedding learning rate accounts for most of μP's advantage over standard parameterization, and a single scaling fix recovers the bulk of the benefit.
Qwen3.6-27B's Dense Architecture Challenges the MoE-Only Playbook for Flagship-Class Coding Models
Alibaba's dense Qwen3.6-27B outperforms its MoE sibling on coding benchmarks, trading predictable inference latency for a larger memory footprint than sparse alternatives.
Chinese AI Models Compared: DeepSeek, Qwen, Kimi, Doubao, and Ernie
Compare DeepSeek, Qwen, Kimi, Doubao, ERNIE and GLM through dated benchmarks, license terms, context windows, API pricing and practical workload fit.
Running DeepSeek R1 Locally: Hardware, Quantization, and Real Throughput
Running DeepSeek R1 locally on real hardware: per-quant VRAM budgets, computed decode ceilings for a 24 GB card, measured quality loss, and sourced throughput figures.
Fish-Speech: The Open-Source TTS Model That's Threatening ElevenLabs
Fish Audio's open-source S2 TTS hit SOTA in March 2026 with sub-100ms latency and 80+ languages, challenging ElevenLabs' commercial pricing and exposing licensing gaps.
Synthetic Data Is Eating AI Training
The supply of high-quality human text for AI training is running out. Synthetic data fills the gap, but model collapse and provenance drift set the limits.
