AI & Machine Learning
ReleasedLaunched

लघुLaghu

Run models bigger than your RAM, on your own machine

Light, small, swift. लघु (Laghu) is a local large-language-model app for Windows, built on Aitele's Srota inference engine: it streams a model's weights from disk under a memory ceiling you set, so models far larger than your free RAM still run on an ordinary PC — trading reply speed for capacity. On a laptop with 1.4 GB free, a 4.4 GB, 7.6-billion-parameter model runs in about 0.8 GB at roughly half a token a second. Its chat interface is served only on your own machine, with no account, API key or upload of your prompts. Version 1.1 adds signed automatic updates that never interrupt a conversation.

Visit Laghulaghu.aiteleresearch.com
Since 2026Co-owned · commons ≤ 40%
§ 01

Research

From the AI & Machine Learning Lab.

  • Technical report · in preparation · NOV 06, 2026
    Srota: Memory-Ceilinged, Disk-Streamed Inference for Running Large Language Models on Commodity PCs

    Running a large language model locally is normally gated by memory: the quantised weights must fit in RAM or VRAM. We describe Srota, the inference engine inside Laghu, Aitele's local LLM application for Windows, which relaxes this constraint by streaming a model's weights from disk under a user-set memory ceiling, so that models larger than available RAM can still be executed — trading generation speed for capacity. Srota reads GGUF containers and decodes the Q4_K, Q6_K, MXFP4, F16, BF16 and F32 block encodings through an independently written portable C kernel that links no ggml code, paired with a byte-level BPE tokeniser implemented from its public description. Version 1.1 hand-vectorises the Q4_K/Q6_K dot products with AVX2, tiles the batched prefill path for cache reuse, and reads streamed blocks ahead on two to four threads inside the ceiling, bypassing the OS file cache when free memory cannot hold the streamed part. On a six-core laptop with an NVMe drive this raised decode speed about threefold over 1.0 at the same ceiling (Qwen2.5 1.5B at 64 MiB: 0.76 → 2.50 tokens/s; Qwen2.5 7B at 256 MiB: 0.15 → 0.44 tokens/s), with word-for-word identical output at temperature 0 and bit-identical logits whether weights are streamed or resident; in the published runs a 7.6-billion-parameter model (4.4 GB) executes in about 0.8 GB at 0.55 tokens/s on a laptop with 1.4 GB free. Laghu ships as a per-user, administrator-free installer that serves a chat interface on the loopback interface only, with signed automatic updates; it requires no account or API key, uploads no prompts and ships no model weights. All measurements come from one machine; characterising the speed–memory-ceiling trade-off across storage media, model sizes and dense versus mixture-of-experts architectures remains the principal open evaluation.

  • Technical report · in preparation · JUL 27, 2026
    Sandhi: A Privacy-Preserving Cross-Device Agent Architecture with Executor-Side Safety Gates

    Agents that operate graphical interfaces usually run in a vendor cloud, see a single device, and rely on the model to police itself. We describe Sandhi, a system that links a user's desktop and Android devices into one end-to-end encrypted network in which the desktop is the only place reasoning happens. At each step a vision-language model — local via Ollama, or user-keyed through an OpenAI-compatible API — receives the target device's accessibility tree (capped at 120 nodes) and a downscaled screenshot, and emits exactly one JSON action from a 26-kind schema. Safety is enforced by the executor, not the model: keyword-classified send, purchase, delete and install actions require confirmation; input into focused password fields is refused; and payment apps cause the agent to disable its own accessibility service. Devices pair through a QR-carried token mixed into an X25519 / HKDF / ChaCha20-Poly1305 handshake, and off-LAN traffic crosses a blind relay that sees only an HMAC-derived room identifier. Licences and update manifests are Ed25519-signed and verified offline, so no account or telemetry is required. Sandhi v0.1.0 ships for Windows and Android. Task success, latency and the robustness of the gates against adversarial models are not yet evaluated; those benchmarks are future work.

  • Preprint · Aitele Research LLP · MAR 15, 2026
    Reverse Synthetic Neural Networks (RSN): Training-Free Model Construction by Closed-Form Statistical Synthesis

    Modern neural networks are trained: a fixed architecture is initialized at random and fit by gradient descent and backpropagation. We study the opposite construction. In a Reverse Synthetic Network the data synthesizes the model directly — every weight, from embeddings and attention projections to feed-forward maps and output heads, is computed in closed form from corpus statistics, with no loss function, no gradient, and no GPU. We report two results, deliberately separated. Synthesis is competitive with trained baselines on discriminative tasks, reaching 87.0% on 20 Newsgroups in tens of seconds on a CPU with zero training. For generation the same paradigm hits a hard ceiling: it collapses to an n-gram model, which we explain mechanistically — PPMI–SVD representations encode distributional similarity, not predictive composition. Measured in bits-per-byte against a locally-run GPT-2, RSN is best understood as a cheap, strong zero-training floor, not a path to the trained frontier.

§ 02

Releases

From the AI & Machine Learning Lab.

  • software · production · NOV 06, 2026
    Srota Inference Engine v1.1

    The disk-streaming LLM inference engine inside Laghu — it streams model weights from disk under a memory ceiling you set, so models larger than your free RAM can still run.

  • software · beta · JUL 27, 2026
    AXON Agent Engine v0.1.0

    The cross-device agent engine inside Sandhi — your desktop and Android devices joined into one end-to-end encrypted network, with the reasoning on your own desktop.

  • software · beta · MAR 15, 2026
    RSN Core v0.4

    Reverse Synthetic Neural Network framework — synthesize full transformer architectures from data in seconds, without training.

§ 03

AI & Machine Learning Lab

What if the data built the network?

A new paradigm for neural architecture — Reverse Synthetic Neural Networks that construct models directly from data patterns, without gradient descent, backpropagation, or GPU training. From zero-training synthesis to interpretable, efficient intelligence.

About the lab →