7 papers
Agaram 2.5B: The First Tamil-First Language Model Trained from Scratch with Tool Calling. A Playbook for Under-Represented Languages
Ramchand Kumaresan, Madhusudhanan G, Kaushik Kezendirane
We present Agaram 2.5B, a Tamil-first language model trained from scratch that converses in Tamil script, romanized Tanglish and English, and calls tools. Our first model, Agaram 1…
Knowing in Advance When an Evolutionary Outer Loop Will Not Help: A Pre-Registered Cheap-Baseline Screening Rule
Ramchand Kumaresan
We introduce a pre-registered screening rule that decides, before any implementation, whether an evolutionary / population / lifecycle outer loop over neural-network parameters or…
Rigel: Reverse-Engineering the Metal 4.1 Tensor Compute Path on the Apple M4 Max GPU
Ramchand Kumaresan
Apple's Metal 4.1 exposes a tensor compute path: the Metal Performance Primitives (MPP) matmul2d operation over cooperative_tensor fragments, whose interface is documented but whos…
Decomposing Evolutionary Mixture-of-LoRA Architectures: The Routing Lever, the Lifecycle Penalty, and a Substrate-Conditional Boundary
Ramchand Kumaresan
We decompose an evolutionary mixture-of-LoRA system on a from-scratch ~150M-parameter widened-D substrate (D=1536, V=32000; D/V approx 0.048; the "widened-1536" substrate) into thr…
KALAVAI: Predicting When Independent Specialist Fusion Works -- A Quantitative Model for Post-Hoc Cooperative LLM Training
Ramchand Kumaresan
Independently trained domain specialists can be fused post-hoc into a single model that outperforms any individual specialist, and the gain is predictable: gain = 0.82 x divergence…
Orion: Characterizing and Programming Apple's Neural Engine for LLM Training and Inference
Ramchand Kumaresan
Over two billion Apple devices ship with a Neural Processing Unit (NPU) - the Apple Neural Engine (ANE) - yet this accelerator remains largely unused for large language model workl…
ACAR: Adaptive Complexity Routing for Multi-Model Ensembles with Auditable Decision Traces
Ramchand Kumaresan
We present ACAR (Adaptive Complexity and Attribution Routing), a measurement framework for studying multi-model orchestration under auditable conditions. ACAR uses self-consistency…