Swissi AI Journal
EN
The Swiss flag flying in front of a snow-covered mountain.

ISSN 3043-1921

Swissi AI Journal Open Access & Peer-Reviewed

Artificial intelligence research from method to deployed system. Editors assess the method, evaluation and evidence, then publish accepted work continuously with open access from its publication date.

How to submitDownload LaTeXOpen in Overleaf

Open access

Peer-Reviewed Swissi Articles

Show all Swissi articles

SAIJ-5kdnql4rsq27Jul 2026

Multi-Jurisdictional Legal Identity Assurance for Capability Gating

A Design-Science Proposal for Tiered, Reusable Identity Assurance of Natural, Juridical, and Machine Entities

DOI 10.5281/zenodo.21901241

Flat maximum verification charges every participant for the rarest high-risk case, and excludes those who cannot clear a bar they never needed to. This model holds the assurance state apart from the capability gate that consumes it, so identity demand follows the act and the weight of its consequences rather than mere presence.

Keywords:
  • identity assurance
  • capability gating
  • tiered and reusable verification
  • multi-jurisdictional identity
  • data minimisation
  • entity taxonomy
  • bitemporal reliance
  • design science research
ReadPDF

SAIJ-zpd6gvtrfaunJul 2026

Credentials and Triangulated Trust Signals on a Single Accountable Identifier

A Hash-Anchored Distributed-Ledger Framework for Portable Identity across Jurisdictions

DOI 10.5281/zenodo.21901243

Digital identity stays rigid while it is bound to provider accounts, mutable handles and local wallet schemes. An accountable hash-anchor tier sits above them: an inert root anchor, unlinkable profile anchors for distinct contexts, and gate-specific assurance evaluated at a point in time.

Keywords:
  • digital identity
  • verifiable credentials
  • accountable pseudonymity
  • distributed ledger
  • selective disclosure
  • identity assurance
  • self-sovereign identity
  • hash anchor
ReadPDF

SAIJ-sh27g6sykt2kJul 2026

Identity-Staked Consensus and Collusion Resistance in Chartered Validator Sets

A Trust Model for Decentralised and Compliant Distributed Settlement Infrastructure

DOI 10.5281/zenodo.21901245

Permissioned ledgers are commonly dismissed as centralised because admission is restricted. Separating permissioning from control distribution makes validator identity externally costly collateral: public legal identity, charter state, liability and audit exposure, with affiliation-aware voting caps and per-member collusion margins.

Keywords:
  • identity-staked consensus
  • permissioned ledger
  • proof-of-authority
  • validator trust model
  • collusion resistance
  • settlement infrastructure
  • actor assurance
  • threshold class coverage
  • ledger evidence record
ReadPDF

SAIJ-f3jtignfyge3May 2026

Firm Valuation When AI Shapes the Business Model

A Milestone-Based Real-Options Framework for the AI Valuation Uncertainty Problem

DOI 10.5281/zenodo.21901247

Discounted cash flow, the IDW S 1 income approach and market multiples compress milestone probabilities, continuation options and risk shifts into opaque aggregate parameters. A milestone-gated real-options overlay decomposes that value into auditable components, with a Success Readiness Index deriving per-option probabilities from structured pairwise comparisons.

Keywords:
  • firm valuation
  • AI integration
  • real options
  • milestone-based valuation
  • intangible assets
  • AHP
  • multi-criteria decision analysis
ReadPDF

SAIJ-cwo7xrcdsautMar 2026

Functional Architecture of European Electricity Trading Markets

Requirements for AI Supported Trading Systems under Regulatory Constraints

DOI 10.5281/zenodo.21901249

European electricity trading runs as a constrained multi-layer system in which legal design, exchange microstructure and network physics execute jointly across forward, day-ahead, intraday and balancing horizons. The paper specifies an AI-supported trading architecture with a permission gate on executable actions and fail-closed control logic under REMIT, MiFID II, MiFIR and EMIR.

Keywords:
  • EU electricity market
  • market coupling
  • NEMO topology
  • electricity balancing
  • AI trading systems
  • compliance-by-design
ReadPDF

SAIJ-xz3bi3q7fwimAug 2025

Compliant AI Infrastructure for Regulated Finance

A tiered multi-agent framework with DLT audit trails for financial operations in DACH

DOI 10.5281/zenodo.21901251

Regulation is treated as an orientation layer rather than a deterministic ruleset: a matrix of regulatory intent and exposure is compiled into concrete prohibitions, obligations and runtime budgets. Evidence, decisions and reason codes bind to a permissioned DAG, so a supervisor can replay how an outcome was reached and attribute failure.

Keywords:
  • DACH finance
  • regulated financial institutions
  • multi agent expert system
  • policy compiled orchestration
  • objective under constraints
  • permissioned DLT
  • DAG timestamping
  • audit trails
  • EU AI Act
  • MiFID II
  • DORA
  • GDPR
  • human oversight
  • execution gating
  • ESG budgets
  • verification and assurance
ReadPDF

SAIJ-ddkjais6s332Aug 2025

A regulatory-compliant AI and verification system for higher education under ESG-aligned constraints

DOI 10.5281/zenodo.21901253

Two linked components for higher education: a role-specific multi-agent framework for institutional operations, and a decentralised verification layer for audit, credential authentication and tamper-evident records. GDPR, the EU AI Act, EQF, ECTS and ESG directives are encoded as structural constraints rather than checked after the fact.

Keywords:
  • Regulatory technology
  • artificial intelligence in education
  • multi-agent AI systems
  • decentralised verification
  • academic tokenisation
  • GDPR compliance
  • EU AI Act
  • digital credential infrastructure
  • ESG governance
  • UniAI
  • UniDVS
ReadPDF

SAIJ-zkihhbpahsbrAug 2025

Verifiable Federated AI Infrastructure

Swiss compliant federated AI DLT network using Nash equilibrium and ESG metrics

DOI 10.5281/zenodo.21901255

Centralised AI infrastructure scales, and collides with latency, auditability and energy constraints. The design separates centralised training from decentralised inference and storage across five node classes, tying a size-neutral availability floor to tiered rewards for service level, ESG performance and anti-concentration.

Keywords:
  • decentralised data centre
  • AI
  • Federated AI infrastructure
  • ESG
  • ESG-aware compute
  • Nash equilibrium
  • digital sovereignty
  • Swiss data regulation
  • tokenised infrastructure
  • verifiable AI services
ReadPDF

SAIJ-soeptiqyucowAug 2025

Generic AI-DLT Enterprise System

Architecture and methodology for scalable domain adaptation from a unified core framework

DOI 10.5281/zenodo.21901257

Compliance in AI deployments is usually applied afterwards, through prompt engineering, rather than built into the foundation. This architecture encodes regulatory, governance and ESG requirements as an objective-under-constraints problem, so every specialised agent operates within legally admissible and auditable bounds before any domain work begins.

Keywords:
  • Compliance-first AI
  • Multi-agent systems
  • Distributed ledger technology
  • Directed acyclic graph
  • Regulation by design
  • ESG integration
  • Domain-agnostic architecture
  • Deployment-agnostic architecture
  • Vendor-agnostic architecture
  • Objective-under-constraints
ReadPDF

SAIJ-qzvrl4bwy7y2May 2025

Multi-Agent AI Architecture for Regulated Insurers

A generic AI framework under Solvency II and the AI Act in Austria and Germany

DOI 10.5281/zenodo.21901259

The insurer is modelled as a constrained optimisation entity under solvency, legal, ESG and operational boundaries, then decomposed into specialised agents for capital, underwriting, claims, compliance and fraud. Human-in-the-loop roles enter through tiered access control, with an orchestrator enforcing regulatory admissibility across the set.

Keywords:
  • Multi-Agent Systems
  • Enterprise AI
  • Insurance Firms
  • Solvency II
  • AI Act
  • Regulated Environments
  • Constrained Optimisation
  • Principal-Agent Theory
  • Nash Equilibrium
  • Arrow’s Risk Pooling
  • Austria
  • Germany
  • Institutional Design
  • ESG Compliance
  • Regulatory Architecture
  • Model Context Protocol (MCP)
  • Agent-to-Agent Protocol (A2A)
  • AI Governance
  • Algorithmic Accountability
  • Financial Regulation
ReadPDF

Standing calls

Current calls for papers

Standing calls identify current editorial priorities within the journal's full scope.

SustainabilityArtificial intelligence and ESG evidence
Results that cost less compute to obtain, artificial intelligence applied to generation and grids, and sustainability reporting treated as a question of evidence.
SovereigntyEnterprise AI under European compliance
Systems that run inside the jurisdiction: self-hosted or European inference, data governed within its required location, and measured capability and financial costs.
LanguageFoundation models for the DACH region
Models trained, adapted and evaluated on German-language material, with benchmarks that exist in German, and a clear account of where a regional model outperforms a general one.

From arXiv

Recent arXiv Articles

arXiv cs.AI

Didactic knowledge or Clinical Cases? How Data Types Shape Medical Large Language Models

Medical large language models are commonly trained on mixtures of didactic data (e.g., textbooks) and clinical data (e.g., patient records), yet how these data types differentially shape model capabilities remains unclear. We address this issue with token-matched experiments that vary the didactic-to-clinical ratio...

arXiv cs.AI2026-09-23

Read at arXiv cs.AI

arXiv cs.LG

"As a Language Model...": Chat Template Switches LLM Self-Referential Voice and Activation Steering Reproduces It

Large Language Models (LLMs) tend to add disclaimers like "I'm just an AI" when asked about something related to themselves. The self-reports from such responses are used in debates about AI safety or self-knowledge of the models, yet what drives them is not well understood. Are the models telling us about...

arXiv cs.LG2026-09-23

Read at arXiv cs.LG

arXiv cs.AI

An Affordable AI-Integrated Smart Cane for Multimodal Mobility Assistance of Visually Impaired Users

Visual impairment affects over 2.2 billion people worldwide, yet conventional white canes cannot detect elevated hazards or provide semantic environmental context. Existing AI-assisted navigation systems typically rely on expensive hardware or cloud connectivity, limiting accessibility in resource-constrained...

arXiv cs.AI2026-09-23

Read at arXiv cs.AI

arXiv cs.LG

Federating Quantum and Classical Computing: A Privacy-Preserving Hybrid Approach

Quantum machine learning (QML) is increasingly recognized as one of the most promising near-term applications of quantum computing, viewed as a next-frontier candidate beyond purely classical approaches. Hybrid quantum-classical models operationalize this potential by embedding a parameterized quantum circuit...

arXiv cs.LG2026-09-23

Read at arXiv cs.LG

arXiv cs.AI

PAANI : On Device Visual Evidence Fusion and Explainable Guidance for River Robot Simulation

Mobile river monitoring robots must interpret obstacles and water boundaries that geographic waypoints alone cannot describe. On resource constrained platforms, converting imperfect visual predictions into timely and inspectable guidance is a distinct challenge. An object label or steering command does not explain...

arXiv cs.AI2026-09-23

Read at arXiv cs.AI

arXiv cs.LG

Entropy Can Flow, or It Can Guide. Be Entropy. LEDFlow: Introducing Entropy-guided Generation Order into Uniform Discrete Flow

Uniform discrete flow permits repeated updates at every generation position. While continued revision supports correction of wrong tokens, it also exposes correct intermediate predictions to later errors. An experiment on Sudoku puzzles shows that 9.4% of generated cells are correct at an intermediate step but...

arXiv cs.LG2026-09-23

Read at arXiv cs.LG

arXiv cs.AI

Social Influence and the Allocation of Scientific Attention in AI Populations

AI systems are becoming participants in the evaluation and use of scientific research. They encounter citation counts, download statistics and lists of popular articles developed around human readers, but the collective consequences of these signals for artificial readers remain uncertain. This paper adapts the...

arXiv cs.AI2026-09-23

Read at arXiv cs.AI

arXiv cs.LG

The Probabilistic Structure of Large Language Models

This paper presents a probabilistic perspective on large language models (LLMs), developed with the aim of bringing together, in a single self-contained account, tools that are usually treated separately across the literature. LLMs are described through probability measures on the set of sequences of tokens,...

arXiv cs.LG2026-09-23

Read at arXiv cs.LG

arXiv cs.AI

Learning 3D biophysical cell properties from 2D images and cell-population statistics

Inferring 3D cellular properties from 2D microscopy is difficult when a reference instrument reports only population statistics rather than labels for individual cells. Here we develop a population-supervised framework that maps single 2D red-cell images to latent biophysical quantities and aggregates them to mean...

arXiv cs.AI2026-09-23

Read at arXiv cs.AI

arXiv cs.LG

Stable Unsupervised Continual Chunking with Sheaf SyncMap

Unsupervised Continual chunking is a fundamental problem in machine learning and neuroscience, where the goal is to identify groups of states that frequently co-occur in temporal sequences. A key challenge is to form accurate chunks while maintaining their stability over time. In this work, we propose sheaf...

arXiv cs.LG2026-09-23

Read at arXiv cs.LG

arXiv cs.AI

Goal-driven Variant Categorization

Process discovery rarely yields a single coherent process structure. For analysis, a common step is to cluster process variants based on structural similarity and then assign business meaning to the resulting groups. Since these partitions are not derived from the organization's goals, analysts must manually...

arXiv cs.AI2026-09-23

Read at arXiv cs.AI

arXiv cs.LG

Brain-Inspired Hierarchical Modularity for General Continual Learning

Continual learning, the ability to learn from sequential experience while retaining and adapting prior knowledge, is central to intelligent systems operating in changing environments. However, conventional continual learning is typically studied with offline task-wise training and clear task boundaries, leaving a...

arXiv cs.LG2026-09-23

Read at arXiv cs.LG

arXiv cs.AI

Replication Without Persistence in Hosted LLMs: Measurement Sensitivity in Action-Time Belief Evaluation

Behavioural evaluations of hosted language models can vary because the evaluated service, the measurement instrument, or both differ across runs. We separate three validation questions: whether a prior finding recurs on fresh data under its historical configuration (replication), whether the endpoint changes when...

arXiv cs.AI2026-09-23

Read at arXiv cs.AI

arXiv cs.LG

Dual-GNN Multilevel Coarsening for Maximum Independent Set

Solving large-scale instances of the Traveling Salesman Problem (TSP) exactly is computationally expensive. Researchers often employ graph sparsification methods to improve computational efficiency. Traditional sparsification methods typically rely on fixed heuristics and fail to fully exploit instance-specific...

arXiv cs.LG2026-09-23

Read at arXiv cs.LG

arXiv cs.AI

The Wisdom of Artificial Deliberative Crowds

The aggregation of many lay estimates often outperforms individual expert judgment, a phenomenon known as the wisdom of crowds. While this is usually attributed to the independence of estimates, an even stronger effect arises through deliberation: averaging the consensus estimates of small deliberating groups...

arXiv cs.AI2026-09-23

Read at arXiv cs.AI

arXiv cs.LG

Exposing Blind Spots in Deep Imbalanced Regression Evaluation

Deep Imbalanced Regression (DIR) addresses a common failure mode of regression models: target distributions are highly non-uniform, causing models to perform best in densely populated target regions even when reliable performance is required across the full target range. Despite rapid methodological progress, DIR...

arXiv cs.LG2026-09-23

Read at arXiv cs.LG

arXiv cs.AI

Agreement Overstates Evidence: Error Dependence in LLM Judge Consensus

Consensus among LLM judges is often taken as strong evidence that a decision is correct. This assumes that judges make their errors independently. In practice, LLM judges are often trained and evaluated in similar ways, so they can make the same mistakes. We study how this dependency affects the reliability of...

arXiv cs.AI2026-09-23

Read at arXiv cs.AI

arXiv cs.LG

Learning Neural Feedback Linearization for Data-driven Systems via Augmented Lagrangian

The paper proposes a novel data-driven framework for designing and training a feedback linearizing controller by explicitly incorporating relative degree based conditions into the learning process. This enables the conventional feedback controller components to be replaced by neural Lie derivatives, thereby...

arXiv cs.LG2026-09-23

Read at arXiv cs.LG

arXiv cs.AI

IntLawNER: A Named Entity Recognition Dataset and Benchmark in International Law

International law provides the normative framework through which states coordinate action, regulate armed conflict, and protect human rights, yet its texts remain without token-level named entity recognition (NER) resources. We introduce IntLawNER, a NER dataset and benchmark for codified sources of international...

arXiv cs.AI2026-09-23

Read at arXiv cs.AI

arXiv cs.LG

Mitigating Sequential Reappearance in Diffusion Data-Point Unlearning

Diffusion data-point unlearning is typically evaluated immediately after each deletion, even though subsequent requests may repeatedly update the same model. We identify sequential reappearance, a failure mode in which an instance that is initially judged to be forgotten later returns to the memorized regime...

arXiv cs.LG2026-09-23

Read at arXiv cs.LG

For authors

Submitting to the journal

Please submit your manuscript, in English or German, as either a LaTeX package or a PDF. Every submission is reviewed by the editors, and every author receives a reasoned response.

Charges
Submission is free. A single CHF 450 charge applies when a manuscript proceeds to external peer review.
Review
Manuscripts selected for peer review are assessed double-blind by at least 2 independent specialists outside the editorial team. Peer review is completed within 4 weeks of submission.

Authors

Research from academia and professional practice

The journal publishes work by researchers, doctoral candidates and practitioners in industry and public institutions. Every manuscript is assessed on its method, evidence and contribution.

Named authors take responsibility for the manuscript and record their individual contributions.

How to submit

Institutions

Collaborative work across laboratories, companies and public bodies

The journal publishes collaborative research conducted across universities, laboratories, companies and public bodies. Evidence from systems in operation is central to its scope.

Named individuals hold authorship; institutions are recorded as affiliations on the published article.

How to submit

Who decides

Editorial board

Editorial decisions are assigned by subject competence. Each manuscript selected for peer review is assessed double-blind by independent specialists. Reviewer identities remain confidential, and the responsible editor makes the final publication decision.

Editor-in-chief

Dr.Walter KurzMBA, M.Sc.

  • Enterprise AI architecture
  • Multi-agent AI systems
  • AI governance and regulatory compliance
  • AI in company valuation and risk
  • Distributed ledger infrastructure
  • Universität Graz
  • Universität Augsburg
  • FH CAMPUS 02
  • Swissi Institute for AI
LinkedIn

Editorial board

Prof. Dr.Velimir Dedić

  • Computer and AI system security
  • Management information systems
  • Data analysis and applied statistics
  • IT security policy
  • E-learning and instructional design
  • FITI Belgrad (Faculty of Information Technology and Engineering)
  • BK University
  • IRRODL
LinkedIn

Editorial board

Prof. Dr.Stefan Marx

  • Internal control systems
  • Statutory audit under HGB and IFRS
  • Corporate governance and compliance systems
  • Internal audit and special audits
  • Sustainability and ESG reporting
  • HfWU Nürtingen-Geislingen
  • Steinbeis-Beratungszentrum Corporate Governance und Wirtschaftsprüfung
LinkedIn

Editorial board

Prof. Dr.Frank Reinhardt

  • Tax law
  • Banking regulation and supervision
  • Corporate governance and internal control
  • Compliance and risk management
  • HfWU Nürtingen-Geislingen
  • Steinbeis-Beratungszentrum Corporate Governance und Wirtschaftsprüfung
LinkedIn

Editorial board

Prof. Dr.Jörg Westphal

  • AI adoption in sales organisations
  • Trust in AI systems
  • AI governance in organisations
  • AI-supported sales training
  • B2B sales management and enablement
  • FOM Hochschule
  • Helmut-Schmidt-Universität Hamburg
  • Universität Hamburg
  • DHBW Stuttgart
  • zfuw Koblenz
  • UCAM Murcia
  • Marketing Management Journal
  • AKAM
LinkedIn

Editorial board

Dr.Ralf Kittelberger

  • Compliance management systems
  • Corporate governance and supervisory board duties
  • Employment law and HR compliance
  • AI in legal work and legal operations
  • Company and commercial law
  • HfWU Nürtingen-Geislingen
  • TU Darmstadt
  • Eberhard-Karls-Universität Tübingen
  • Fortbildungsinstitut der Rechtsanwaltskammer Stuttgart
LinkedIn

Editorial board

Dr.Florian Kollberg

  • AI risk management in regulated finance
  • Model risk and validation
  • Financial regulation and FINMA requirements
  • ESG reporting and sustainable supply chains
  • Corporate finance, M&A and company valuation
  • HfWU Nürtingen-Geislingen
LinkedIn

Editorial board

Dr.Wojtek StrickerM.Sc.

  • AI in company valuation
  • Corporate finance and investment cases
  • KPI and controlling systems
  • Group steering and strategic controlling
  • Renewable energy finance
  • Signum Magnum College
  • FernUniversität in Hagen
  • Frankfurt School of Finance & Management
  • Corporate Finance Institute
LinkedIn

Editorial board

Prof. Dr.Svetlana Andjelic

  • Computer adaptive testing
  • Assessment design and measurement
  • Database systems and data modelling
  • Software architecture and domain-driven design
  • Education technology
  • Singidunum University Belgrade
  • Union University
  • FON, University of Belgrade
  • ITS Belgrade
  • Faculty of Computer Science Banja Luka

Editorial board

Prof. Dr.Nenad Dedić

  • Applied cryptography
  • Cloud and data security
  • Software supply-chain security
  • Secure AI architecture and deployment
  • Distributed systems
  • Boston University
LinkedIn

Editorial board

Prof. Dr.José Machado

  • Industrial automation and robotics
  • Control systems and mechatronics
  • Industrial systems modelling and simulation
  • Industrial digitalisation and Industry 4.0
  • Manufacturing systems engineering
  • University of Minho (MEtRICs Research Center)
  • École Normale Supérieure de Cachan

Editorial board

Prof. Dr.Šemsudin Plojović

  • Applied statistics
  • Management information systems
  • Knowledge management and e-business
  • Innovation ecosystems and entrepreneurship
  • University of Novi Pazar

Editorial board

Prof. Dr.Ivica Stankovic

  • Quantitative risk modelling
  • Model risk management and validation
  • AI governance in financial institutions
  • Market and credit risk analytics
  • University College Dublin
  • University of Belgrade
LinkedIn

Editorial board

Prof. Dr.Enes Sukic

  • Information systems and service architecture
  • Scholarly publishing and peer review
  • Research methodology and reporting
  • Management of technology
  • FITI Belgrad (Univ. Union-Nikola Tesla)
  • University of Niš
  • University of the Balearic Islands
LinkedIn

For readers and libraries

Terms of publication

These publication terms apply to every article and form the journal record used by libraries, indexes and funders.

ISSN
3043-1921
Key title
Swissi AI journal
Abbreviated key title
Swissi AI j.
Publisher
Swissi Holding AG, Zug, Switzerland
Editor-in-chief
Dr. Walter Kurz, MBA, M.Sc.
Publication model
Continuous publication. Each article receives an article number.
Review
Manuscripts selected for peer review are assessed double-blind by at least 2 independent specialists from outside the editorial team. Peer review is completed within 4 weeks of submission. Review is double-blind: author and reviewer identities are withheld from each other.
Access
Open access. Full text is available from publication.
Charges
Submission and reading are free. A single CHF 450 charge applies when a manuscript proceeds to external peer review.
Licence
Creative Commons Attribution 4.0 International. Authors keep copyright.
Preservation
Every article is deposited with Zenodo, operated by CERN.
Listed in
ROAD, the ISSN International Centre's directory of open access scholarly resources. The ISSN Portal record is public in full, and the ISSN was assigned by the ISSN Centre Switzerland at the Swiss National Library.
Language
Articles are published in English with a German HTML reading version. The English text is the version of record, and each article PDF is published in English. Manuscripts are accepted in English or German.

Editorial standard

What the editors read for

Every submission is assessed against six criteria, including reproductions and negative results.

MethodReproducible from what is written
The method identifies the data and provenance, model, training and inference settings, and the decisions needed to reproduce the work.
EvaluationBaselines, conditions and failures
The evaluation reports baselines, ablations, operating conditions, failures and the setup behind each central result.
ClaimsConclusions stay within the evidence
A benchmark result is reported as a benchmark result, and a claim about behaviour in service rests on measurements from service.
Prior workWhat it stands on, and what it adds
The article states the literature it stands on and what it adds to it. Where a finding contradicts published work, it engages with that work directly.
Systems in serviceDeployed systems, as they behave
Reports on deployed systems state operating conditions, observed behaviour, measured effects and failures.
Reproduction and negative resultsFull editorial weight under the stated criteria
Reproductions, refutations and negative results carry full editorial weight when the method and evaluation meet the journal's criteria.

Scientific integrity

The journal applies these standards.

Kodex Wissenschaftliche Integrität

Swissuniversities, the Swiss National Science Foundation and Innosuisse drew up a code of conduct for scientific integrity together, under the lead of the Swiss Academies of Arts and Sciences.

PDF (DE)akademien-schweiz.ch

1.00 MB · 4 May 2021

© 2021 Akademien der Wissenschaften Schweiz. Open-access publication under CC BY 4.0, source doi.org/10.5281/zenodo.4707584.

Recent Daily Papers

From Hugging Face

Hugging Face

Hugging Face Daily Papers

RoboFollow: Unveiling the Instruction Following Mirage in Embodied Agents

Modern embodied agents achieve impressive success rates, yet their actual instruction-following ability is far weaker than these numbers suggest. We trace this illusion to a structural property we term low scene entropy: when a visual scene admits only one valid task, language becomes redundant and a policy can...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

Blaming Across the Aisle: Political Contrasting and Blame Attribution in the Danish Parliament

Political discourse is widely perceived to be growing more hostile, yet robust evidence remains scarce. This study examines blame attribution in the Danish Parliament from 1997 to 2026, combining a purpose-built classifier, BlameBERT (F1: 0.80), with multilevel statistical modeling. The classifier is constructed...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

ImIR: Image-Instruction Tuning for All-in-One Image Restoration

Degradations vary widely across images, so a practical restoration system has to handle many degradation types with one model. A recent and effective recipe adapts a large pretrained image-editing model to restoration using a small low-rank adapter with a text prompt. We replace that prompt with an instruction...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

Emergent Collusion in Long-Horizon LLM Agent Interaction

LLM agents are increasingly deployed in collaborative settings, yet long-term interaction may give rise to undesirable coordination. We study the emergence of collusion in a long-horizon multi-agent environment: two agents repeatedly complete individual tasks, share task logs, verify each other's work, and receive...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

Lean Pool: An AI-Maintained Archive of Formalized Mathematics

Lean Pool is a repository of formalized mathematics. It is grown, maintained and optimized by AI agents.

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

The Tasteful Agent: Measuring and Improving Taste in Long-Horizon Tasks

LLM agents increasingly work on long-horizon tasks, and the decisions they make along the way, such as which hypothesis to test or which implementation to build on, determine the outcome of the whole run. Making these decisions well is becoming a key capability for both engineering and research agents. We refer to...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

StableVQ: Practical Guidelines for Stable Vector-Quantized Tokenizer Training

Vector Quantization (VQ) is fundamental to discrete visual tokenizers that power modern autoregressive and masked image generation models. While recent shared-projection codebook methods have substantially advanced codebook utilization, training stability remains a critical and underexplored challenge. We argue...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

ALPINE: Adaptive Localization for Parameter- and Sample-Efficient Few-Shot Learning

Few-shot learning research is predominantly evaluated on accuracy alone, with limited attention to the parameter and training-sample budgets required to reach that accuracy - a real constraint for practitioners without large-scale compute. We present an ultra-lightweight (22,249-34,917 parameter) spatial-relational...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

All-in-One Multilingual Scene Text Recognition with Script-aware Mixture-of-Experts

Multilingual scene text recognition (STR) remains challenging due to the scarcity of training data for most languages and the difficulty of serving diverse scripts within a single model. Existing solutions either deploy one recognizer per language, inflating cost and introducing error accumulation, or rely on...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

Geometric and Semantic Coupling for Interaction Understanding in 3D Scenes

Interaction understanding in 3D scenes requires a joint description of movable parts, their motion, and the regions through which they can be operated. We present Segment-Snap, which connects these outputs through the physical relationship between parts and handles. Learned predictors identify broad part surfaces...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

Circuit Hypernetworks for Quantum-Augmented Diffusion Language Models

Language models can be adapted by changing the computations applied to individual tokens. Quantum circuits offer one such approach, but evaluating wider circuits inside a large model can be computationally demanding. Here we introduce HyperQ, which adds token-conditioned quantum residual branches to a frozen...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

GAE: Learning a Geometry-Native Latent Space for 3D-Consistent World Generation

We present a compact geometry-native latent space as a shared foundation for perception and generation. Visual generators can produce photorealistic frames without preserving a consistent 3D scene. We argue that this is not only a modeling problem but also a representation problem: generators typically evolve...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

Bellman Policy Optimization

Reinforcement learning with verifiable rewards (RLVR) improves the reasoning capabilities of large language models (LLMs). We introduce Bellman Policy Optimization (BPO), a critic-free method derived from Policy Mirror Descent (PMD). For autoregressive generation with terminal rewards, BPO uses the Bellman...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

From Pattern Recognizers to Personalized Companions: A Survey of Large Language Models in Mental Health

The rising global prevalence of mental health conditions, together with longstanding barriers in traditional healthcare, such as limited resources, high cost, stigma, and privacy concerns, has created an urgent need for accessible and scalable support. Large Language Models (LLMs) have emerged as a transformative...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

Ovis-Embedding: Pushing the Frontiers of Universal Omni-Modal Embeddings

In this report, we introduce Ovis-Embedding, a state-of-the-art omni-modal embedding family built on native integration of text, image, video, and audio. Instead of assembling separate modality towers, Ovis-Embedding uses a shared multimodal backbone to encode different modalities in a common representation space....

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

Recursive self-improvement of AI research agents

AI agents are beginning to automate research and development across the AI stack, from improving training efficiency to optimizing inference. A natural next step is to improve the research efficiency of the agents themselves. When an AI research agent's own code is the object of optimization, each accepted rewrite...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMs

Diffusion Large Language Models (dLLMs) have recently emerged as a promising alternative to autoregressive LLMs by enabling non-autoregressive text generation. However, their practical deployment remains limited by inefficient inference, largely due to the absence of effective Key-Value (KV) caching and scalable...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

RULER: Instance-aware Rubric Rewards for SVG Generation

Generating Scalable Vector Graphics (SVG) code from natural-language instructions is an open-ended task without absolute visual ground truth, leaving both evaluation and policy optimization without a faithful signal. Scalar metrics (CLIP, Aesthetic) calibrated on natural images transfer poorly to stylized vector...

Hugging Face2026-09-23

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

The information geometry of large language models is shared, learned, and controllable

Large language models learn similar behaviours, yet it remains unclear what structure they share or how to change one behaviour without disturbing others. The Fisher-Rao geometry of next-token probabilities connects these questions: behaviour determines this geometry up to output-preserving symmetries, whereas...

Hugging Face2026-09-22

Read at Hugging Face
Hugging Face

Hugging Face Daily Papers

Towards Full Pipeline FP8 Reinforcement Learning for LLMs

Reinforcement learning (RL) has become a key technique for improving the reasoning and agentic abilities of large language models (LLMs). Although FP8 quantization can accelerate RL training, maintaining stability throughout an FP8 RL pipeline remains challenging. While previous works have focused on resolving...

Hugging Face2026-09-22

Read at Hugging Face

This site stores functional cookies for language, consent and regional routing, plus the theme in browser storage. Choose your cookie setting. Privacy policy