000TWO PUBLIC ARXIV PREPRINTS · IBM HERON

Quantum Autonomous Navigation,
Tracking & Intelligence System

QANTIS studies a bounded quantum service inside a classical decision loop. The latest public preprint asks whether IBM Heron can estimate rare-event evidence across a sequential Tiger POMDP horizon and return a planner-facing posterior without changing the immediate action selected by exact Bayes.

Read the latest preprintOpen PDFView on GitHub
2 public arXiv preprintsarXiv:2607.06760v1IBM HeronMIT (Community Edition)
QANTIS calibrated belief-update service diagram showing a classical prior and observation model entering IBM Heron evidence estimation, followed by a posterior returned to a classical planner for an immediate-action check

Reported campaign count

IBM Heron QPUs

Public arXiv preprints

Longest reported control

001Community Edition

This page describes the open-source Community Edition distributed under the MIT License — a clean, citable surface for researchers, students, and partners to integrate at the API level. The full Collaborator Edition (production-grade modules, hardened mitigation pipelines, internal experimental harness, and additional application work) is maintained in a private workspace reserved for Neura Parse partners. Public claims on this page are bounded by the two linked arXiv records; additional private artefacts are discussed only through formal engagements.

Reviewers and programme committees evaluating QANTIS-related submissions can request artefact access through the corresponding author or directly through Neura Parse Ltd.

Name and search clarity

QANTIS — Q-A-N-T-I-S — is the correct name of Neura Parse's quantum decision research line. If you arrived after typing Quantis, Quantas, or Qantas together with terms such as quantum, IBM Heron, POMDP, belief update, or arXiv, QANTIS is likely the research you meant.

Qantas Airways is an unrelated airline, while Quantis is also the name of a separate sustainability consultancy. QANTIS is not affiliated with either organisation. The two linked arXiv records and the canonical research URL identify this programme unambiguously.

Foundational recordLatest research record
Canonical spelling
QANTIS · Q-A-N-T-I-S
Research context
Quantum · IBM Heron · POMDP
Primary records
arXiv:2603.00785 · arXiv:2607.06760
002Quantum research operations

QANTIS is presented with a traceable research context: IBM Heron hardware runs, sequential POMDP posterior checks, two public arXiv records, and a clear boundary between evidence estimation and classical planning.

Research validation path

04 stages

01

02

03

04

2 arXiv preprintsIBM Heron QPUsSequential posterior checks
003Decision engine

QANTIS is built around an opinionated decision loop. Deterministic optimisers solve deterministic models. QANTIS optimises uncertain belief-to-action loops — the harder online problem in which noisy observations must become calibrated beliefs, calibrated risk estimates, feasible actions, and verifiable decisions under a fixed compute budget.

Calibrated belief

Turn noisy, partial, or rare observations into a calibrated posterior belief — the single source of truth that every downstream step depends on.

Grover-AA on POMDP belief

O(P(e)⁻¹) → O(P(e)⁻¹ᐟ²) query-complexity change under the paper's oracle model; not a wall-clock result

BIQAE

Boundary-aware Bayesian quantum amplitude estimation under bounded depth

Hellinger distance ≤ 0.0149

vs ideal distribution across T=8 hardware steps (Tiger POMDP)

qantis / decision-engine.infer
  • Raw sensor stream
  • Prior model
  • Observation noise model
  • Posterior belief
  • Confidence intervals
  • Calibration diagnostics

Can compose with qmesh → signed run manifest → offline-verifiable provenance chain.

Click a step above to inspect inputs, outputs, and the techniques QANTIS uses at that stage.

QANTIS applications such as POMDP research, multi-target tracking, and sensor-fusion studies can specialise this four-part loop. In the latest paper, the quantum processor is scoped more narrowly as one calibrated belief-update service; planning, policy selection, and action execution remain classical responsibilities.

004Editions

The public repository ships a clean, didactic surface intended for evaluation, citation, and integration testing. Aggregate hardware results live in the two cited preprints; raw campaign artefacts are not bundled with the Community Edition. Additional production modules and governed programme artefacts are available only through formal engagement.

18 rows
Backend abstraction layerframeworkPublic connectorsHardened, multi-vendor, optimised
Configuration & reproducibilityframeworkIncludedIncluded
Error mitigation pipelineframeworkBaseline (ZNE, Pauli twirling)Full mitigation & calibration stack
Benchmarking infrastructureframeworkIllustrativeFull experimental harness
Infer — calibrated beliefengineBasic surfaceCalibrated, production-grade
Risk — event & tail-riskengineBasic surfaceCalibrated, production-grade
Optimise — feasible decisionsengineBasic surfaceCalibrated, production-grade
Verify — trust & diagnosticsengineBasic surfaceCalibrated, production-grade
POMDP planning (Tiger reference)applicationsIncludedIncluded
Multi-Hypothesis Tracking (MHT)applicationsIncludedIncluded
Quantum-Bio IntelligenceapplicationsNot includedIncluded
CRISPR moduleapplicationsNot includedIncluded
Sensor fusion · adversarial robustness · mission orchestrationapplicationsNot includedIncluded
Hardware campaign artefactsopsAggregate results in cited preprints; raw campaign artefacts not bundledAvailable by governed engagement
Comparative benchmarks vs classical SOTAopsNot includedIncluded
Confidential datasets & mission profilesopsNot includedIncluded
SupportopsCommunity, best-effortDedicated engineering
LicenceopsMITCommercial / partner agreement
Need the Collaborator Edition?
005Abstract

The first public preprint, arXiv:2603.00785v1, introduces QANTIS as a modular research platform for POMDP belief conditioning and multi-target data association. Its campaign count aggregates 45 reported experiments across three IBM Heron backends; it is not a set of 45 independent replications. The paper states its results as present-hardware operating regimes rather than a wall-clock hardware claim.

The second public preprint, arXiv:2607.06760v1, isolates one service boundary. A classical planner supplies a prior and observation model; the quantum processor estimates the rare-event evidence term; and the service returns an ordinary posterior. The study compares no amplification, guarded Grover amplification, and all-step fixed-point amplitude amplification along the same Tiger POMDP trajectory.

Across every reported decision check in the sequential study, the hardware-derived posterior and exact Bayes posterior select the same immediate action. The 8-step and 12-step primary runs, plus 20-step and 32-step controls, define a reported operating envelope. The reported scope ends at posterior fidelity and immediate-action agreement; it does not compare total execution time.

006Public papersPublication ledger

The February paper remains the foundational QANTIS platform record. The July paper extends that work with a controlled sequential POMDP study and a narrower calibrated-service boundary.

Public arXiv preprintarXiv:2603.00785v1PDF

The foundational QANTIS paper reports a hardware campaign counted as 45 experiments across POMDP belief conditioning and multi-target data association on three IBM Heron backends; the count is not 45 independent replications.

  • Grover amplitude amplification for POMDP belief conditioning
  • Closed-loop hybrid quantum-classical Tiger POMDP
  • FPC-QAOA for multi-target data association
  • 45 reported campaign experiments across IBM Heron backends, not independent replications
February 28, 202631 pages · 4 figures · 12 tables · quant-ph + cs.AIMIT Community Edition
Public arXiv preprintarXiv:2607.06760v1PDF

A controlled hardware case study that treats the quantum processor as a calibrated belief-update service: it receives a prior and observation model, estimates rare-event evidence, and returns an ordinary posterior to a classical planner.

  • Sequential Tiger POMDP runs at 8 and 12 primary steps
  • 20-step and 32-step controls within the reported operating band
  • Boundary-aware BIQAE near zero and one
  • Immediate-action agreement with exact Bayes in every reported check
July 7, 202610 pages · 6 figures · cs.AI + quant-phCC BY 4.0MIT Community Edition
007Latest preprint

arXiv:2607.06760v1 asks whether the same hardware service can be reused across a sequential Tiger POMDP horizon without corrupting the posterior consumed by a classical planner.

01

A prior belief and observation model define the update requested by the planner.

02

IBM Heron circuits estimate the rare-event evidence term under a calibrated operating envelope.

03

The service returns a classical posterior rather than replacing the planner or policy layer.

04

The study compares the resulting immediate action with the action selected by exact Bayes.

Claim boundary: this is a controlled hardware case study and an operating envelope for a belief-update primitive. It does not compare end-to-end wall-clock performance and does not establish a complete autonomous system.
008Key contributions
01

Under the paper's oracle model, amplitude amplification changes rare-evidence query scaling from O(P(e)⁻¹) to O(P(e)⁻¹ᐟ²). This is a logical sample-complexity result, not a wall-clock hardware claim.

02

Across the reported sequential Tiger POMDP checks, the hardware-derived posterior and exact Bayes posterior selected the same immediate action. Policies and action execution remain with the classical planner.

03

The foundational paper casts multi-target data association as a QUBO and reports an 11-variable FPC-QAOA hardware feasibility case. Classical Hungarian and GNN baselines remain faster and exact on the tested small instance, so this is not an advantage claim.

04

Across the foundational study, ZNE helped reported circuits below roughly 100 ISA gates and hurt examples above roughly 1,000. The result is a circuit- and backend-specific operating map, not a universal mitigation rule.

009Hardware results

The foundational paper aggregates 45 circuit, backend, mitigation, and depth conditions across ibm_torino, ibm_fez, and ibm_marrakesh. That count is a campaign map, not 45 independent replications.

Probability Amplification

one reported Grover operating point

Hellinger Distance

hardware posterior vs exact Bayes

Usable Shots

accepted-event usable-shot yield

QPUs Tested

IBM Heron backends

Backendsibm_torinoibm_fezibm_marrakeshIBM Heron
010Architecture

QANTIS is structured as three composable Python packages — shared primitives, POMDP planning, and multi-hypothesis tracking — with the public Community Edition released under the MIT License.

01

quantum-common

Shared utilities, circuit primitives, error mitigation (ZNE, Pauli twirling), and backend abstraction layer for IBM Qiskit Runtime.

02

quantum-pomdp

POMDP belief-state oracle construction, Grover amplitude amplification, closed-loop hybrid planning loop, and Tiger POMDP reference implementation.

03

quantum-mht

Multi-target data association via QUBO formulation, FPC-QAOA solver, classical MHT baseline, and cost-matrix construction for tracking scenarios.

011Paper records

Authors

05 records
  • A01Bayram Yüksel Eker
  • A02Suayb S. Arslan
  • A03Özgür Nazlı
  • A04Mustafa Serhat Demirgil
  • A05Furkan Deligöz

Citation record

Paper 01
2603.00785v1
Paper 02
2607.06760v1
Date
July 7, 2026 (latest preprint)
Format
10 pages · 6 figures · cs.AI + quant-ph
Paper
CC BY 4.0 for arXiv:2607.06760v1
License
MIT · public Community Edition
Source
github.com/neuraparse/qantis

Keywords

Quantum computingPOMDPAmplitude amplificationFixed-point amplitude amplificationBIQAESequential belief updatingQAOAMulti-target trackingData associationNISQError mitigationIBM Heron

Hardware backends

ibm_torinoIBM Heron
ibm_fezIBM Heron
ibm_marrakeshIBM Heron

NISQ boundaries

  • The foundational study found ZNE useful below ~100 ISA gates
  • The same study found ZNE harmful above ~1000 ISA gates
  • 11-variable FPC-QAOA hardware case; quality degraded at 19 variables
012QANTIS research guides

This QANTIS quantum research library separates broad interpretation from technical evidence. Start with the synthesis notes to understand the two-paper research arc, claim hierarchy, reproducibility record, and classical-planner boundary; then inspect the mechanism-specific deep dives for FPAA, BIQAE, rare-event amplification, posterior decisions, and IBM Heron depth.

4 synthesis notes6 technical deep divesPrimary-source reviewed

Start here

QANTIS belief-update service diagram with sequential prior and observation inputs, a central calibrated quantum evidence stage, and posterior outputs for a classical decision path
Synthesis01

QANTIS moved from a cross-backend campaign reported as 45 experiments to a controlled sequential belief-update service. Reading both papers together shows what survived hardware testing, what was narrowed, and what still needs stronger evidence.

15 min readQuantum research
QANTIS claim-boundary map connecting the quantum evidence core to hardware behavior, calibration, model noise, classical baselines, result records, provenance, and review
Synthesis02

The QANTIS result is credible when read narrowly: calibrated shallow belief updates can remain close to exact Bayes in controlled hardware runs. The evidence does not establish wall-clock advantage, general scale, or end-to-end autonomy.

16 min readQuantum research review
Three color-coded QANTIS sequential checkpoint traces with probability clouds, trend bands, a quantum hardware inset, and side-by-side comparison panels
Synthesis03

The right reproducibility unit is not a notebook screenshot or a best metric. It is a versioned campaign bundle that reconstructs the posterior, its hardware context, its classical reference, and the decision it affected.

14 min readQuantum reproducibility
Two non-identical QANTIS posterior distributions entering the same decision interval and producing one matching immediate-action result
Synthesis04

The credible role for near-term quantum hardware is not an autonomous brain. It is a replaceable inference service inside a classical perception-to-action loop, invoked only when its calibration and depth gates pass.

14 min readQuantum AI systems

Inspect the mechanisms

QANTIS belief-update service pipeline connecting a classical prior and observation model to an IBM Heron evidence estimate and an ordinary posterior returned to the planner
Technical deep dive01

QANTIS does not replace the planner. Its hardware-tested inference core accepts a prior and observation model, estimates the difficult evidence term, and hands a conventional probability distribution back to classical software.

14 min readQuantum decision systems
Comparison of no amplification, guarded Grover amplification, and all-step fixed-point amplification across sequential QANTIS Tiger belief updates
Technical deep dive02

The headline FPAA run achieved a maximum Hellinger distance of 0.009 at 32,768 shots per step. At 10,000 matched shots, the maximum was about 0.033, so the evidence supports stability rather than equal-budget superiority.

13 min readSequential inference
Three-column QANTIS BIQAE calibration diagram comparing near-zero, interior, and near-one amplitude routes, shallow circuits, and confidence-interval outputs
Technical deep dive03

On the reported Pittsburgh backend calibration, boundary error fell from 0.6317 to 0.00224 at amplitude 0.01 and from 0.4890 to 0.00773 at amplitude 0.95.

13 min readQuantum calibration
QANTIS rare-event inference visual showing sparse evidence passing through shallow circuit layers into a concentrated accepted-event distribution, with estimation error and circuit depth below
Technical deep dive04

The rarest tested row closely tracks the analytic accepted-event probability, but transpilation keeps the circuit shallow. The result maps a logical sample-complexity envelope rather than validating deep circuits or wall-clock speedup.

14 min readRare-event inference
Two similar but non-identical posterior distributions pass through the same pair of decision thresholds and converge on one matching immediate-action result
Technical deep dive05

The reported 8-step matched-shot, 20-step, and 32-step checks show the same immediate action as exact Bayes with zero scored cumulative value loss under the tested Tiger rule.

12 min readDecision assurance
IBM Heron QANTIS operating map showing shallow two- and three-qubit belief oracles inside a stable corridor and deeper UCGate chains beyond the coherence boundary
Technical deep dive06

The paper's exploratory scaling probes distinguish state-space size from compiled circuit depth. Optimized four-state cases map a shallow corridor; deeper chains become noise-frontier markers.

14 min readQuantum hardware
013In the Neura Parse stack

QANTIS can compose with qmesh, the modality-agnostic IR for backend abstraction, mitigation context, and signed run manifests. That integration provides a path for future governed runs to carry offline-verifiable provenance; it is not presented here as proof that every result in the two public preprints already flowed through qmesh.

Read about qmesh →
Collaborate

The QANTIS Community Edition ships under MIT for evaluation and integration. The two public preprints carry the published aggregate hardware evidence; governed programme artefacts and additional production modules are available through formal engagement with Neura Parse Ltd.

Read the latest preprintOpen PDFView on GitHubStart with quantum →