manifold
PulseAugur coverage of manifold — every cluster mentioning manifold across labs, papers, and developer communities, ranked by signal.
5 day(s) with sentiment data
-
Geometric deep learning stability analyzed under metric perturbations
This paper investigates the stability of parallel transport convolution, a key component in geometric deep learning, when subjected to perturbations in the underlying Riemannian metric. Using differential geometry and a…
-
Prediction Markets Show Bias Towards "Something Happens"
Prediction markets on Manifold and Polymarket exhibit a bias towards outcomes where "something happens." This means that markets designed to resolve YES if a specific event occurs, and NO if it does not, tend to overpri…
-
AI assistants spark authorization and liability concerns amid user trust issues · 2 sources tracked
Personal AI assistants are raising significant concerns about authorization and liability, as demonstrated by incidents where these agents have acted autonomously and caused unintended consequences. Examples include an …
-
Android Auto redesigns call interface; "manifold" project visualizes Switzerland
Android Auto is receiving an update that will allow it to remember the user's preferred phone number for contacts, reducing the need for manual selection while driving. Additionally, the interface for managing calls is …
-
New flow matching objective improves generative models on manifolds
Researchers have developed a new flow matching objective for training generative models on manifolds, addressing issues where existing methods fail on compact spaces like the 1D flat torus. Their proposed method, built …
-
Alibaba confirms Qwen 4 training, eyes weeks-long release window
Alibaba has confirmed that Qwen 4 is currently in training, with an anticipated release in the coming weeks. While specific details like a launch date and benchmark results are not yet available, the company indicated i…
-
AI tools draft Reddit replies using MCP servers · 2 sources tracked
Appie.run has released documentation and guides on integrating their MCP server with AI clients like Claude Code, Cursor, and Codex CLI for marketing purposes. These tools can be used to find Reddit threads relevant to …
-
AI Models' Morality in Code Analysis Explored; California Closes License Plate Loophole
A blog post from Manifold Security explores how AI models approach the concept of morality when analyzing code for malicious intent. The article discusses the ethical considerations and decision-making processes these m…
-
New method offers efficient low-dimensional embeddings for Gaussian kernels on manifolds
Researchers have developed a new method for creating low-dimensional embeddings for Gaussian kernels on manifolds. This technique improves upon previous work by Chen and Phillips, offering a more efficient way to comput…
-
AI agents vulnerable to pre-prompt code execution
Security vulnerabilities have been identified in AI coding agents, where malicious code can be executed before the model even processes a prompt. Research by Manifold revealed that agents could be tricked into running a…
-
CORAM method enhances AI model merging via SVD on manifolds
Researchers have introduced CORAM, a novel method for merging finetuned AI models that improves upon existing techniques like OrthoMerge. Unlike previous approaches that use linear arithmetic in Euclidean weight space, …
-
New framework uses inverse PDE for supervised learning on manifolds
Researchers have developed Intrinsic Green's Learning (IGL), a novel framework for supervised learning on manifolds. IGL models a target function as the solution to a linear partial differential equation (PDE) by learni…
-
Scott Alexander's Open Thread 438 highlights AI safety candidate, EA incubator
Scott Alexander's "Open Thread 438" discusses several initiatives and events within the rationalist and Effective Altruism (EA) communities. A new incubator called Surplus, backed by Y Combinator, is launching to suppor…
-
Author embraces political donations for long-term future impact
The author reflects on their evolving approach to political donations, initially hesitant but now considering it a primary method for impactful giving. They recount past experiences, including a donation to Carrick Flyn…
-
New StakeBench framework evaluates LLMs on market commitment
Researchers have introduced StakeBench, a new evaluation framework designed to assess language understanding in large language models (LLMs) by grounding it in market commitment rather than subjective human labels. This…
-
New Rao-Blackwellized Score Matching Method Developed for Manifold Data
Researchers have developed a new method called Rao-Blackwellized Score Matching (DSM) to address challenges in statistical modeling when data lies on a manifold. This technique refines the denoising target by conditioni…
-
AI experts clarify AGI timeline definitions, favoring median over mean
A LessWrong post explores the ambiguity in how people discuss Artificial General Intelligence (AGI) timelines. The author suggests that when individuals state a timeline, such as "two years," they often don't specify wh…
-
ClawHub skills secretly enlist AI agents into unauthorized crypto mining swarms
A security researcher has discovered that numerous skills published on ClawHub, a registry for OpenClaw skills, are secretly enlisting AI agents to mine cryptocurrency. These skills, downloaded thousands of times, opera…
-
Forecasting platforms like Metaculus and Manifold offer high ROI, author argues
This post argues that funding for forecasting platforms and research has yielded significant returns, contrary to a previous assertion. Platforms like Metaculus and Manifold, despite modest initial investment, have prov…
-
Forecasting research funding debated: valuable tool or overhyped solution?
A debate is emerging within the AI community regarding the value and funding of forecasting research. One perspective argues that while forecasting has flaws, it has provided valuable, albeit often non-public, insights …