talk-data.com

Showing 2 results

Activities & events

Title & Speakers	Event
CAIML 38 - Agentic GraphRAG with MCP servers 2025-09-09 · 16:30 Register: https://www.meetup.com/de-DE/cologne-ai-and-machine-learning-meetup/events/305829303/ We will have two talks with additional time for networking. Talk 1: Tomaz Bratanic (Graph ML and GenAI research at Neo4j): Agentic GraphRAG with MCP servers Talk 2: Pablo Iyu Guerrero (AI Inference Engineer at Aleph Alpha) and Lukas Blübaum (AI Engineer at Aleph Alpha): Tokenizer-free language model inference	CAIML 38 - Agentic GraphRAG with MCP servers
Tokenizer-free language model inference 2025-05-20 · 19:00 A team from Aleph Alpha will talk about tokenizer-free language model inference. This talk presents an approach to language model inference that eliminates the need for conventional large-vocabulary tokenizers, using a core vocabulary of 256 byte values and a three-part architecture (byte-level encoder/decoder, a latent transformer, and patch embeddings). The talk will cover the architecture and engineering challenges in building an efficient inference pipeline, coordinating models, CUDA graphs, and KV caches. byte-level encoder/decoder patch embeddings latent transformer cuda graphs kv caches inference pipeline	Tokenizer-free language model inference

Title & Speakers

Event

CAIML 38 - Agentic GraphRAG with MCP servers 2025-09-09 · 16:30

We will have two talks with additional time for networking. Talk 1: Tomaz Bratanic (Graph ML and GenAI research at Neo4j): Agentic GraphRAG with MCP servers Talk 2: Pablo Iyu Guerrero (AI Inference Engineer at Aleph Alpha) and Lukas Blübaum (AI Engineer at Aleph Alpha): Tokenizer-free language model inference

CAIML 38 - Agentic GraphRAG with MCP servers

Tokenizer-free language model inference 2025-05-20 · 19:00

A team from Aleph Alpha will talk about tokenizer-free language model inference. This talk presents an approach to language model inference that eliminates the need for conventional large-vocabulary tokenizers, using a core vocabulary of 256 byte values and a three-part architecture (byte-level encoder/decoder, a latent transformer, and patch embeddings). The talk will cover the architecture and engineering challenges in building an efficient inference pipeline, coordinating models, CUDA graphs, and KV caches.

byte-level encoder/decoder patch embeddings latent transformer cuda graphs kv caches inference pipeline

Tokenizer-free language model inference

Showing 2 results