Inference + Moe | hypedar

Inference + Moe

57.0

Create an automated pipeline for fine-tuning Mixture-of-Experts models on consumer-grade hardware. This enables local, high-reasoning capabilities without expensive cloud inference.

+411

activeimplementation gap

moequantizationinference

Signals (31)

oobabooga/textgen

From Dispersion to Attraction: Spectral Dynamics of Hallucination Across Whisper Model Scales

CSAttention: Centroid-Scoring Attention for Accelerating LLM Inference

Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7

Identity as Attractor: Geometric Evidence for Persistent Agent Architecture in LLM Activation Space

Geometric Routing Enables Causal Expert Control in Mixture of Experts

Demonstration of Pneuma-Seeker: Agentic System for Reifying and Fulfilling Information Needs on Tabular Data

diegosouzapw/OmniRoute

nvidia blog1d ago

Rethinking AI TCO: Why Cost per Token Is the Only Metric That Matters

Neural networks for Text-to-Speech evaluation

The local LLM ecosystem doesn’t need Ollama

Pushing the Limits of On-Device Streaming ASR: A Compact, High-Accuracy English Model for Low-Latency Inference

Pro Max 5x quota exhausted in 1.5 hours despite moderate usage

google/magika

Small models also found the vulnerabilities that Mythos found

WAND: Windowed Attention and Knowledge Distillation for Efficient Autoregressive Text-to-Speech Models

Leveraging Weighted Syntactic and Semantic Context Assessment Summary (wSSAS) Towards Text Categorization Using LLMs

HUOZIIME: An On-Device LLM-enhanced Input Method for Deep Personalization

yt-dlp/yt-dlp

Response-Aware User Memory Selection for LLM Personalization

Cloudflare's AI Platform: an inference layer designed for agents

jamiepine/voicebox

MemGround: Long-Term Memory Evaluation Kit for Large Language Models in Gamified Scenarios

When Self-Reference Fails to Close: Matrix-Level Dynamics in Large Language Models

Awakening Dormant Experts:Counterfactual Routing to Mitigate MoE Hallucinations

The beginning of scarcity in AI

obsidianmd/obsidian-clipper

Bouncer: Block "crypto", "rage politics", and more from your X feed using AI

Qwen3.6-35B-A3B: Agentic coding power, now open to all

Temperature-Dependent Performance of Prompting Strategies in Extended Reasoning Large Language Models

Equifinality in Mixture of Experts: Routing Topology Does Not Determine Language Modeling Quality