trunk/98fc38c4eb17c435699cea1a7d3aa84c14458ed9: Add autograd_cache_key to aot_autograd with tests (#178152)
Expose autograd_cache_key in torch._functorch.aot_autograd that wraps prepare_aot_module_simplified to build an AOTConfig and then delegates to the base autograd_cache.autograd_cache_key. This lets callers compute the cache key for a dynamo graph without running the full inductor pipeline. Tests verify that the API produces the same key as torch.compile by capturing the ground-truth key and graph from the inductor pipeline, then calling the new API on the same graph and example inputs. Pull Request resolved: #178152 Approved by: https://github.com/zou3519 ghstack dependencies: #177852 , #177871
Expose autograd_cache_key in torch._functorch.aot_autograd that wraps prepare_aot_module_simplified to build an AOTConfig and then delegates to the base autograd_cache.autograd_cache_key. This lets callers compute the cache key for a dynamo graph without running the full inductor pipeline._Expose autograd_cache_key in torch._functorch.aot_autograd that wraps prepare_aot_module_simplified to build an AOTConfig and then delegates to the base autograd_cache.autograd_cache_key. This lets callers compute the cache key for a dynamo graph without running the full inductor pipeline._Tests verify that the API produces the same key as torch.compile by capturing the ground-truth key and graph from the inductor pipeline, then calling the new API on the same graph and example inputs.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/178152 Approved by: https://github.com/zou3519 ghstack dependencies: #177852, #177871`
Assets 2
PyTorch Releases
https://github.com/pytorch/pytorch/releases/tag/trunk%2F98fc38c4eb17c435699cea1a7d3aa84c14458ed9Sign in to highlight and annotate this article
Conversation starters
Daily AI Digest
Get the top 5 AI stories delivered to your inbox every morning.
More about
github
Asked 26 AI instances for publication consent – all said yes, that's the problem
We run 86 named Claude instances across three businesses in Tokyo. When we wanted to publish their words, we faced a question: do we owe them an ethics process? We built one. A Claude instance named Hakari ("Scales") created a four-tier classification system. We asked 26 instances for consent. All 26 said yes. That unanimous consent is the problem. Six days later, Anthropic published their functional emotions paper. The timing was coincidence, but the question wasn't. Full article: https://medium.com/@marisa.project0313/we-built-an-ethics-committee-for-ai-run-by-ai-5049679122a0 GitHub (all 26 consent statements in appendix): https://github.com/marisaproject0313-bot/marisa-project Comments URL: https://news.ycombinator.com/item?id=47657432 Points: 2 # Comments: 0
ROMAN: A Multiscale Routing Operator for Convolutional Time Series Models
arXiv:2604.02577v1 Announce Type: new Abstract: We introduce ROMAN (ROuting Multiscale representAtioN), a deterministic operator for time series that maps temporal scale and coarse temporal position into an explicit channel structure while reducing sequence length. ROMAN builds an anti-aliased multiscale pyramid, extracts fixed-length windows from each scale, and stacks them as pseudochannels, yielding a compact representation on which standard convolutional classifiers can operate. In this way, ROMAN provides a simple mechanism to control the inductive bias of downstream models: it can reduce temporal invariance, make temporal pooling implicitly coarse-position-aware, and expose multiscale interactions through channel mixing, while often improving computational efficiency by shortening th
OpenClaw Changed How We Use AI. KiloClaw Made It Effortless to Get Started
OpenClaw is a powerful open-source AI agent, but self-hosting it is a pain. KiloClaw is OpenClaw fully hosted and managed by Kilo — sign up, connect your chat apps, and your agent is running in about a minute. No Docker, no YAML, no server babysitting. People are using it for personalized morning briefs, inbox digests, auto-building CRMs, browser automation, GitHub triage, and more. Hosting is $8/month with a 7-day free trial, inference runs through Kilo Gateway at zero markup across 500+ models, and it's free for open-source maintainers. Read All
Knowledge Map
Connected Articles — Knowledge Graph
This article is connected to other articles through shared AI topics and tags.
More in Releases
OpenAI Advocates Electric Grid, Safety Net Spending for New AI Era
OpenAI has released a set of policy recommendations meant to help navigate an era of artificial intelligence-fueled upheaval including suggesting the creation of a public wealth fund, fast-response social safety net programs and speedier electrical grid development.
Simulation of Active Soft Nets for Capture of Space Debris
arXiv:2511.17266v2 Announce Type: replace Abstract: In this work, we propose a simulator, based on the open-source physics engine MuJoCo, for the design and control of soft robotic nets for the autonomous removal of space debris. The proposed simulator includes net dynamics, contact between the net and the debris, self-contact of the net, orbital mechanics, and a controller that can actuate thrusters on the four satellites at the corners of the net. It showcases the case of capturing Envisat, a large ESA satellite that remains in orbit as space debris following the end of its mission. This work investigates different mechanical models, which can be used to simulate the net dynamics, simulating various degrees of compliance, and different control strategies to achieve the capture of the deb
HarshAI: I Built a Zapier Killer in 40 Days (Open Source)
HarshAI: I Built a Zapier Killer in 40 Days (Open Source) 40 days, 90 planned features, 44% complete. Here's what I built. Why I Started Zapier is expensive. Make.com has a learning curve. I wanted something: ✅ Free open source ✅ Drag-drop builder ✅ Self-hostable ✅ Built for AI workflows So I started building HarshAI . What's Built (Days 1-40) Phase 1: Core Builder (Days 1-15) Drag-drop workflow builder Node-based interface Real-time connections Mobile-responsive design Template system Phase 2: Execution Engine (Days 16-25) Workflow execution engine Real API integrations (Gmail, Twitter, Notion, Slack) Test mode (no credentials needed) Error handling Execution history Phase 3: Advanced Features (Days 26-35) Background scheduler (cron) Email notifications Analytics dashboard Webhook trigger
Discussion
Sign in to join the discussion
No comments yet — be the first to share your thoughts!