Maya-Sutra-P1: Letting Go on Silicon Characterizing the Vairāgya Forgetting Signature on BrainScaleS-2
Abstract
Version 2 (2026-08-04). Revised to the NLL Universal Paper Format v6. The original text is retained in full; nothing has been deleted. Corrections appear as marked blocks placed at the section that carries the claim, and each one states what the paper said, what the data show, the corrected claim, why it happened, and what still stands. This version adds apparatus and two scoping lines; no claim is retracted.The paper claims only what it measured, discloses that the decay is applied in software rather than on-chip, and correctly disclaims mechanism novelty against Pehle et al. (2022) and Cramer et al. (2022).Two notes: five consecutive days cannot bound drift, as the paper states; and the Vairāgya framing is a label rather than a mechanism, which the paper's own scope note concedes. Chip access route added to Data availability. Everything below this line is the original description from version 1. It is retained unchanged for the record. Where it conflicts with the corrections above, the corrections stand. Two framings it repeats have since been withdrawn in full — the Bhaya Quiescence Law and the Buddhi S-Curve. Both are addressed in Maya-Meta P1, now superseded, and in the self-audit of Maya-Meta P2. Maya-Sutra P1: Letting Go on Silicon characterizes, on real analog silicon (BrainScaleS-2), how a single neuron stops firing ("lets go") as its input synaptic weight is reduced. Using a minimal in-the-loop protocol — set a weight, deliver a fixed five-spike burst, count the resulting spikes — we report five properties of this Vairāgya (non-attachment; the forgetting mechanism) signature. (1) The decay curve is reproducible within a session (maximum spread 0.47 spikes across three identical runs). (2) The shape is rate-invariant in weight-space; an apparent rate-dependence in round-space is a sampling artifact. (3) The weight-to-firing response is a soft staircase whose analog noise is localized at a let-go edge, with near-noiseless plateaus. (4) Across five consecutive days on independent nightly calibrations the shape reproduces while the absolute edge drifts within a bounded band (weights 37–39, mean 38.2). (5) Across a 32-neuron sample the staircase is the majority behaviour (~78% once the measurement window is widened), with neuron-to-neuron fixed-pattern variation dominating day-to-day calibration drift. The underlying analog mechanisms — soft firing thresholds, fixed-pattern heterogeneity, calibration drift — are known BrainScaleS-2 physics; no mechanism-level novelty is claimed. The contribution is a clean, honest characterization, the Vairāgya lens with predictions the hardware can falsify, and the neuron-reliability groundwork required before a property confirmed on digital substrates can be credibly tested on analog silicon. Two artifacts were caught and rejected, not published: a rate-space "linger" that proved to be a sampling artifact, and two stale-file uploads detected by a timestamp-and-noise check. Scope-lock per Maya-Meta P2: all findings are reproducible signatures within THIS architecture under THIS protocol — Property, not Law; a characterization and reliability baseline, not new physics and not a claim about biological brains or all chips. Series: Part of the Maya-Sutra Series — Nexus Learning Labs' hardware-native track, characterizing neuromorphic substrates directly (one substrate, one honest characterization at a time) as the foundation for cross-substrate confirmation of the Maya program's properties. Research characterization only. Links: GitHub Repository (private — to request access: email research@nexuslearninglabs.in with subject "Code Access Request — Maya-Sutra-P1" and your research context) | Interactive Dashboard | FAQ | Full Series Index — nexuslearninglabs.in Nexus Learning Labs, Bengaluru · UDYAM-KR-02-0122422 · BHASKAR IN-0526-9452JS · ORCID: 0000-0002-3315-7907 · VAIRAGYA_DECAY_RATE = 0.002315 — an ORCID-derived provenance mark, not an experimental parameter. Each paper's disclosure block states whether it reached a result in that paper. · Canary: MayaNexusVS2026NLL_Bengaluru
