convaiinnovations
/

laya

Text Classification

calibrated-decisions

reinforcement-learning

Model card Files Files and versions

Resources

View closed (1)

Higher-Context

#17 opened about 5 hours ago by

Free hosted endpoint for Laya, Jev-compatible API

#16 opened about 12 hours ago by

Zero-shot Laya as an agent step guard: agrees when agents say "done", never escalates destructive commands, blind past 320–768 tokens

#15 opened 1 day ago by

Reproduced your 78.33% on MASSIVE exactly, but struggling on my own data. What am I missing?

#14 opened 1 day ago by

A small real-world calibration check on Laya typed-decisions

#13 opened 1 day ago by

CPU latency numbers for laya-multilingual — the model card only has T4

#12 opened 1 day ago by

Real view of convaiinnovations/laya

#11 opened 1 day ago by

Decision model benchmarks

#10 opened 1 day ago by

Catalog selection: low gold-label accuracy and candidate-order sensitivity with Laya-MLX (multilingual included)

#9 opened 2 days ago by

First-run evaluation: K-12 school email triage (what worked, what didn't)

#8 opened 2 days ago by

Laya feels like an embedding/similarity matcher, not a context-understanding model (benchmark included)

#7 opened 2 days ago by

Independent Laya vs Jev head-to-head on identical inputs

#6 opened 3 days ago by

Laya-MLX Metal Memory Growth Fix

#5 opened 3 days ago by

What was the dataset used?

#4 opened 3 days ago by

Laya-multilingual: a 100-scenario decision-making evaluation

#3 opened 4 days ago by

choice:11+ temperature (0.1006) saturates confidence: answers return at 1.00, wrong ones included

#2 opened 5 days ago by