45 stories tagged with #parameter, in publish-time order across the WeSearch catalog. Tag pages update as new stories ingest.
⌘ RSS feed for this tag → or search "Parameter"
Running a 28.9M parameter LLM on an $8 microcontroller
Comment on How to Automate Hyperparameter Optimization by king99
A step-by-step guide into performing a hyperparameter optimization task on a deep learning model by employing Bayesian Optimization that uses the Gaussian Process. We used the gp_m…
Tencent releases Hy3, a 295B-parameter model that it says is competitive with GLM-5.1 and 5.2, under the Apache 2.0 license, following a preview launch in April (Sam Witteveen/VentureBeat)
Sam Witteveen / VentureBeat : Tencent releases Hy3, a 295B-parameter model that it says is competitive with GLM-5.1 and 5.2, under the Apache 2.0 license, following a preview launc…
Meituan open-sources LongCat-2.0, a 1.6T-parameter model that it says was trained on a 50K-chip cluster of domestic Chinese processors, without giving details (Reuters)
Reuters : Meituan open-sources LongCat-2.0, a 1.6T-parameter model that it says was trained on a 50K-chip cluster of domestic Chinese processors, without giving details — China's f…
Scaling MoE Models with LongCat-2.0: A Deep Dive into 1.6T Parameter Architecture Design
Explore LongCat-2.0's 1.6T parameter MoE architecture and its breakthroughs in scalability, efficiency, and performance for next-gen AI systems.…
Grounded Iterative Language Planning: How Parameterized World Models Reduce Hallucination Propagation in LLM Agents
arXiv:2606.27806v1 Announce Type: new Abstract: World models for language agents come in two useful forms. An agent-based world model calls an LLM API and reasons flexibly in langu…
NVIDIA drops DGX Station for Windows (1-Trillion Parameter desktop). Who else is ready to run LLaMA-Behemoth locally?
Nvidia unveils DGX Station, a desktop Windows PC powered by its GB300 Grace Blackwell chip with up to 748 GB of memory, capable of running 1T-parameter models (Mike Wheatley/SiliconANGLE)
Mike Wheatley / SiliconANGLE : Nvidia unveils DGX Station, a desktop Windows PC powered by its GB300 Grace Blackwell chip with up to 748 GB of memory, capable of running 1T-paramet…
Nvidia says RTX Spark offers up to 20 CPU cores and a Blackwell GPU with 6,144 CUDA cores, capable of "100 FPS 1440p gaming" or running 120B-parameter models (Jeffrey Kampman/Tom's Hardware)
Jeffrey Kampman / Tom's Hardware : Nvidia says RTX Spark offers up to 20 CPU cores and a Blackwell GPU with 6,144 CUDA cores, capable of “100 FPS 1440p gaming” or running 120B-para…
Parallax: Parameterized Local Linear Attention for Language Modeling
SupraLabs 50M Parameter Model Just Hit the Trending Page on Hugging Face 🤯
Fitting the parameters of a Besace curve like the Meta logo
The Meta logo resembles a Besace curve. How to find the parameters of a Besace curve to fit an image, such as the Meta logo.…
Pitfalls of Estimating Parameters from Aggregates
One of the most common mistakes in data analysis is treating observed aggregates as if they are the parameters themselves. Observed data is not the parameter — it is the result of …
Training freezes during PSO hyperparameter search
Does the Responses API store parameter save on input tokens?
Poverty Bayes: fitting million-parameter models for pennies with serverless MCMC
Fitting a million-parameter Bayesian model on Modal with PyMC and NumPyro…
0G trains 107B parameter decentralized model with China Mobile, a first for AI above 100 billion parameters
0G Labs trained a 107B-parameter AI model with China Mobile using its DiLoCoX framework, achieving 357x communication efficiency over standard 1 Gbps links.…
SkillOpt treats markdown skill files as trainable parameters with proper optimization machinery
AI Metrics Decoded: From Parameters to TOPS
AI Metrics Decoded: The Numbers That Actually Matter in Production Why You...…
Geo-Expert: Towards Expert-Level Geological Reasoning via Parameter-Efficient Fine-Tuning
While general-purpose Large Language Models (LLMs) applied to Geology often hallucinate when reasoning about subsurface structures and deep-time evolution, current AI in Earth scie…
DCGAN inference on a microcontroller: 12.6M parameters, 512KB SRAM, 26-second generation, pure C [P]
Small Model, Big Brain: The 27B Parameter Model Is the New King of Code
Discover why Qwen3.6-27B is a game-changer. Learn how this dense model brings flagship coding power to local hardware, changing the AI landscape for good.…
Designing XSLT transforms with parameters and multiple inputs
Many real-world transformations do not run on a single XML document. You often merge a primary...…
When Good Equations Get Bad Scores: Improving Symbolic Regression Through Better Parameter Optimization
Symbolic Regression (SR) plays a central role in scientific knowledge discovery by distilling mathematical equations from observational data. Most existing SR methods function with…
How do ML practitioners select hyperparameters, architectures, etc for self-supervised representation learning when the loss is non-monotonic? [D]
Why not dynamic active parameters (and other questions for the knowledgeable)
768GB of cheap Intel Optane DIMM memory sticks used to run 1-trillion-parameter LLM on a system with a single GPU — local Kimi K2.5 install achieved roughly 4 tokens per second
768GB of cheap Intel Optane DIMM memory sticks used to run 1-trillion-parameter LLM on a system with a single GPU — local Kimi K2.5 install achieved roughly 4 tokens per second
Redditor found 768GB of affordable Optane sticks second-hand.…
The Return of Recursion: How 5M-Parameter Models Are Outperforming Frontier LLMs on Reasoning in 2026
The Return of Recursion: How 5M-Parameter Models Are Outperforming Frontier LLMs on...…
A 0.12% parameter add-on gives AI agents the working memory RAG can't
Cohere releases Command A+, a sparse MoE open model built for agentic tasks, with 218B total and 25B active parameters, its first under the Apache 2.0 license (Carl Franzen/VentureBeat)
Carl Franzen / VentureBeat : Cohere releases Command A+, a sparse MoE open model built for agentic tasks, with 218B total and 25B active parameters, its first under the Apache 2.0 …
Cerebras says its chips run a trillion-parameter AI model nearly 7 times faster than GPU clouds
Cerebras Brings Trillion Parameter Inference to Enterprises with Kimi K2.6
Cerebras brings Kimi K2.6 to enterprises, delivering near-1,000 TPS inference on the first trillion-parameter open-weight model for real-time AI coding.…
A Nonlinear Complexity Index for Wearable PPG Cardiovascular Stability: Multiscale Validation, Systematic Evaluation Correction, and Bayesian Parameter Optimization
Cardiovascular stability estimation from wearable photoplethysmography (PPG) requires a principled nonlinear framework, yet major gaps persist in heuristic parameter selection and …
Optimize_anything: A Universal API for Optimizing Any Text Parameter
Can a single LLM-based optimization system match specialized tools across fundamentally different domains? We show that when optimization problems are formulated as improving a tex…
I built a coding agent that gets 87% on benchmarks with a 4B parameter model, here's how
GQA-{\mu}P: The maximal parameterization update for grouped query attention
Hyperparameter transfer across model architectures dramatically reduces the amount of compute necessary for tuning large language models (LLMs). The maximal update parameterization…
Agentic LLM Inference Parameters Reference for Qwen and Gemma
This page is a practical reference for agentic LLM inference tuning (temperature, top_p, top_k,...…
Kristen Stewart's "I just don’t think that it’s possible to create sort of radical, vital work under capitalistic parameters."
Nous Research Releases Token Superposition Training to Speed Up LLM Pre-Training by Up to 2.5x Across 270M to 10B Parameter Models
Nous Research Releases Token Superposition Training to Speed Up LLM Pre-Training by Up to 2.5x Across 270M to 10B Parameter Models
Climate Change crisis: IMD to change parameters to declare heatwaves
Climate Change crisis: IMD to change parameters to declare heatwaves | India News…
US startup Poolside debuts its first open-weight model, Laguna XS.2, a 33B-A3B-parameter MoE model, and Laguna M.1, a proprietary 225B-A23B-parameter MoE model (Carl Franzen/VentureBeat)
Carl Franzen / VentureBeat : US startup Poolside debuts its first open-weight model, Laguna XS.2, a 33B-A3B-parameter MoE model, and Laguna M.1, a proprietary 225B-A23B-parameter M…
Consequences of passing too few register parameters to a C function
It's bad news no matter how you slice it, but Itanium makes it even worse.…
The Conspiracy Against High Temperature Sampling
The Conspiracy Against High Temperature LLM Sampling - fk_top_p_and_top_k.md…