--- title: "AI engineering, November 2, 2026" description: "The Forward Pass daily issue, November 2, 2026." canonical: https://theforwardpass.net/archive/daily/2026-11-02 updated: 2026-11-02 --- # AI engineering, November 2, 2026 The Forward Pass daily issue, November 2, 2026. Source: https://theforwardpass.net/archive/daily/2026-11-02 > This issue is researched and written by AI models, and every fact is checked against its cited source. No human edits it before it is sent. _Text JEPA hits 64% top-5 accuracy, Samsung forecasts 2027 capacity crunch, Mitsubishi Heavy plans $490 million turbine expansion, SenseTime releases SenseNova-Loop-9B-Preview_ Samsung’s contract chipmaking business spent years struggling to fill its factories. Now Samsung expects demand to exceed capacity at every node below 8nm by 2027. Meanwhile, researchers report 64% top-5 accuracy on ImageNet-100 in a pair-free comparison, versus at most 56% for the cited language models up to 8B parameters. Mitsubishi Heavy Industries plans a $490 million Georgia factory expansion to increase U.S. gas-turbine manufacturing and maintenance capacity. If you read one thing today: the JEPA paper. Use its DINOv2 and CLIP alignment results to compare a 109M-parameter text encoder against evaluated language models up to 72B parameters. ## Text JEPA hits 64% top-5 accuracy in pair-free ImageNet-100 test **Top News** · 124 HN points This paper tests a text JEPA for aligning separately trained text and vision encoders. It reports 64% top-5 accuracy on ImageNet-100 in a pair-free zero-shot comparison, versus at most 56% for cited language-model results up to 8B parameters. Here is what makes this credible: - The 109M-parameter text JEPA was trained on 900M image-derived text samples. - The authors report better alignment with DINOv2 and CLIP vision encoders than every language model evaluated, including models up to 72B parameters. - Separately trained JEPAs can learn representations that agree up to a rotation when their positive pairs are independent observations of the same world state. - Captioning combines a text-trained JEPA and causal decoder with a frozen vision encoder, using Gaussian noise to handle residual alignment error. One catch: the captioning setup fits its rotation using a small set of image-caption pairs. The pair-free result applies to the separate benchmark, and the alignment result does not establish a general-purpose replacement for language models. Sources: [hyper.ai](https://hyper.ai/en/papers/2610.39302) ## Samsung projects a 2027 capacity crunch across every node below 8nm **Top News** · Reported · digitimes.com Samsung’s contract chipmaking business spent years struggling to fill its factories. Now Samsung says demand will exceed capacity across every node below 8nm by 2027, a reversal from that earlier problem. Here’s what changed: - The third-quarter earnings-call forecast covers all nodes below 8nm, not just a single process node. - Utilization at those nodes is expected to exceed 100%, putting projected demand above available capacity. - The analysis attributes the projected crunch to demand from Samsung’s own memory unit. One catch: this is Samsung’s forecast for 2027, not a report that capacity is already exhausted. Why care? Samsung’s own memory unit is behind the projected capacity crunch, according to the analysis. The forecast concerns all nodes below 8nm. Sources: [digitimes.com](https://www.digitimes.com/news/a20261029VL222/samsung-business-2027-capacity-revenue.html) ## Mitsubishi Heavy plans $490 million expansion for U.S. gas-turbine capacity **Top News** · Reported · nikkei.com Mitsubishi Heavy Industries plans to expand its Georgia factory to serve the rush to build data centers. The $490 million investment will increase U.S. gas-turbine manufacturing and maintenance capacity. Here’s what the investment covers: - The Georgia factory expansion is intended to serve the rush to build data centers. - More manufacturing capacity will go toward gas turbines, a key component in power plants that generate electricity using natural gas. - Maintenance capacity will grow alongside manufacturing, so the expansion is not limited to producing new equipment. Why care? This is a planned $490 million investment in equipment used by natural-gas power plants. For infrastructure engineers, it is an investment in gas-turbine manufacturing and maintenance capacity to track alongside the compute buildout. Sources: [asia.nikkei.com](https://asia.nikkei.com/business/energy/mitsubishi-heavy-gears-up-to-boost-us-gas-turbine-production) ## Signals 1. [SenseTime releases SenseNova-Loop-9B-Preview, an open-source model running 56 layer passes from 32 layers](https://pandaily.com/sensetime-sensenova-loop-9b-preview-looped-transformer-mhc-open-source-apache-2) 2. [nv-uefi-unlocker boosts FP16 performance by up to 65% on a developer’s laptop RTX 4060](https://www.tomshardware.com/pc-components/gpus/open-source-usb-tool-unlocks-nvidias-half-speed-gpu-ai-limiter-on-windows-lifts-gpus-fp16-math-performance-up-to-65-percent-uefi-unlocker-covers-rtx-20-30-and-40-series-desktops-and-laptops) 3. [LOWN decoder retains 99.6% of LLaDA-V's score using 35.3% of decoding FLOPs on average](https://hyper.ai/en/papers/2610.36671) 4. [SPaRC benchmark shows redrawing puzzles improves VLM accuracy more than changing prompts](https://hyper.ai/en/stories/364a5c6b8245b912c44aff5c0732f12b) 5. [L3G-World generates spatially consistent infinite worlds at 9 FPS using latent 3D memory](https://hyper.ai/en/papers/2610.40614) 6. [T3 Code controls Claude Code and Codex from mobile, web and desktop apps](https://github.com/pingdotgg/t3code) · 159 Stars today