PAPER PLAINE

Fresh research, simply explained. Updates twice daily.

Europe's Climate Ambition Under Scrutiny: Evidence from Deep Learning Emission Projections

Europe will miss its climate targets by a third without urgent policy changes

Europe's current trajectory will overshoot its 2030 climate target by 35%, according to machine learning projections of emissions across all EU countries through 2023. While renewable energy is cutting power sector emissions as planned, transportation has barely improved and now accounts for over one-third of total emissions, revealing a gap that spans nearly all member states rather than concentrating in a few laggards.

The EU has staked its global climate credibility on the 55% reduction target, but hitting it now requires far more aggressive intervention than current policies deliver. Transportation's structural slowness—affecting countries across the bloc—means the problem can't be solved by a few policy fixes; Europe needs wholesale changes to how people and goods move, not just incremental tweaks to existing measures.

DMT-Dens: Density-preserving manifold visualization for biological data

A better way to visualize rare cells without distorting the crowded ones

Scientists created DMT-Dens, a visualization tool that displays high-dimensional biological data while accurately preserving density patterns — keeping crowded regions visibly dense and sparse regions visibly sparse. Unlike existing methods, it prevents the visual distortion that can make rare or transitional cell states hard to spot or misinterpret their actual frequency in a sample.

Cell biologists rely on these visualizations to understand which cell types are common, which are rare, and which represent transitions between states — critical for discovering new cell populations and understanding disease progression. Distorted density patterns have led researchers to misidentify the prevalence of rare cells, potentially steering biological discovery in the wrong direction. DMT-Dens lets biologists see the true distribution of their samples, making it easier to catch genuinely rare events and understand the actual landscape of cell diversity.

Policy-Invariant Reward Shaping from LLM Feedback: A Framework for Hybrid RL Agents

Why imperfect AI feedback won't ruin learning robots

When large language models guide reinforcement learning agents, the AI feedback doesn't need to be perfect to work. Researchers proved mathematically that even inaccurate language model scores preserve the optimal strategy an agent learns, and tested this claim on systems where the misleading feedback was twenty times stronger than the true signal.

Building AI systems that combine language models with learning agents is becoming standard practice, but engineers haven't had theoretical assurance that imperfect feedback won't poison the results. This work provides that assurance, letting teams use language model guidance without needing to validate every single score—saving engineering time on systems where language model feedback is cheaper and faster than manually annotated data.

The Zonotopic Mixture Filter

Tracking systems that work with both probability and hard boundaries

Engineers have created a new way to estimate the location and state of moving objects that combines two previously separate approaches: probabilistic methods that use probability distributions, and guaranteed-bounds methods that promise hard limits on error. The zonotopic mixture filter splits the difference by treating noise as coming from one of several bounded sets chosen at random, then uses multiple parallel tracking systems weighted by their probability of being correct. The method guarantees it will give the right answer a specified percentage of the time while still respecting hard physical bounds.

State estimation is critical in robotics, autonomous vehicles, power grids, and aerospace—anywhere you need to know what's happening in a system based on noisy sensor data. This approach gives engineers certainty about worst-case scenarios (like aircraft safety margins) without ignoring the statistical patterns in real-world noise, combining safety guarantees with realistic performance. For systems where both statistical accuracy and hard worst-case bounds matter, this eliminates the need to choose between two competing frameworks.

Where A Small Language Model Helps in Invoice Categorisation, Understood Through Embedding Geometry

How a small AI model learns to sort invoices—and what geometry reveals about why

A small language model trained on a single computer reached 96% accuracy at automatically sorting invoices into the correct accounting categories, outperforming larger zero-shot AI systems. The researchers discovered that the model's success hinges on geometric patterns in how it represents financial language—clusters of similar invoices that map closely to vendor identity—and found a surprising mismatch: structured invoice formats that help human accountants actually make the AI model perform worse.

Companies increasingly automate invoice sorting for tax compliance and financial reporting, but relying on large cloud-based AI creates cost and security risks. This work shows that smaller, in-house models can match or beat expensive alternatives while keeping sensitive financial data private—and only need about 100 labeled invoices per new client to work well. Understanding *how* these models make decisions through geometry rather than black-box predictions also lets accountants spot when something has gone wrong.

From Corpora to Co-Evolving Capabilities: Capability-Centric Data Design for Generalist Image Generation

Building image generators by teaching skills in the right order

Researchers built a new system for training image-generation AI that organizes training data around learning dependencies—teaching fundamental skills before advanced ones—rather than treating each task separately. The approach created a 440-million-image dataset and produced models that handle text-to-image generation and image editing with broader visual coverage and better skill transfer than conventional methods.

Image generators trained this way perform better across multiple tasks without needing separate specialized models, reducing computational waste and engineering overhead. This suggests that how data is sequenced during training matters as much as the data itself—a finding that could improve efficiency in training other large AI systems.

The geometric Laplace transform: Definition, existence and properties of the Geometric Algebra Laplace transform

Extending a classic math tool to work with geometric algebra

Researchers have defined the Laplace transform—a fundamental mathematical tool for solving differential equations—to work within geometric algebra, a system that handles multidimensional spaces and rotations. This extension applies to geometric algebras with signature 5 or lower, filling a gap needed for modeling and controlling real-world systems like electrical circuits.

Engineers designing electrical circuits and other dynamical systems increasingly want to use geometric algebra because it naturally represents rotations and multidimensional transformations. Without a properly defined Laplace transform in this framework, they have to convert between different mathematical systems, losing efficiency and clarity. This work removes that barrier, letting engineers stay within geometric algebra from problem setup through solution.

Diff-DDoS: Realistic Cyber-Physical Attack Synthesis and Robust Detection for 5G-Enabled CPS Using Tabular Diffusion Models

Making DDoS attack detectors work with realistic threats instead of fake ones

Current AI systems that catch DDoS attacks on 5G networks fail badly when facing real attacks, losing 47 to 100 percent accuracy. Researchers built a new tool using diffusion models to create realistic fake attacks, then used those to train detectors until they stayed accurate against the real thing—recovering 79 to 100 percent accuracy depending on attack type.

5G networks power critical infrastructure like hospitals and power grids. Today's DDoS detectors are brittle and collapse when attackers adapt their methods slightly, creating serious security gaps. This technique lets companies test and harden their defenses against realistic attacks before deployment, rather than discovering failures during an actual breach.

Quipu: A Governed Bitemporal Knowledge Graph Store

A database that keeps AI agents honest about what they write

When AI agents write data to knowledge graphs, existing databases use outdated rules designed for human curators: they accept writes first and check them later, trust all writers equally, and keep minimal history. Quipu inverts this approach, forcing every write through a governance gate before acceptance, tracking complete time histories with trust labels, and treating governance rules themselves as queryable facts—catching all six planted defects in testing versus zero catches in ungated systems.

As AI systems increasingly generate and share information autonomously, the stakes of unchecked writes grow. Quipu makes governance verifiable through audit trails that live inside the database itself, so organizations can answer "Did this agent follow our rules?" as a simple query rather than manual forensics. Testing shows it answers all real governance questions correctly while reducing false claims by up to 87.5% compared to baseline approaches.

Is the medium the message? Social disclosure channels and firm risk

Where companies announce social issues shapes how investors view risk

Companies that disclose social and labor issues for the first time through SEC filings see their stock risk increase, while ongoing disclosures in sustainability reports lower risk. The channel matters as much as the message: unexpected information published in SEC filings signals genuine news, whereas routine updates in dedicated sustainability reports reassure investors.

Investors trade billions based on how they perceive company risk, and this research shows that the same social disclosure can either alarm or comfort investors depending on where it appears. Companies planning to disclose social problems face a real choice: burying new concerns in standard SEC filings will likely spook markets, while establishing regular sustainability reporting and integrating disclosure into financial reports can actually reduce perceived risk. Regulators and investors should treat first-time SEC social disclosures as material signals worth attention.

A Neurofinance Framework for Subjective Temporal Perception, Risk, and Investment Behavior

How your brain's valuation shifts your sense of financial time

The way your brain evaluates investments physically reshapes your subjective experience of time—meaning two financially identical choices can feel different lengths depending on neural activity. Brain imaging of 1,183 investment decisions revealed that when neural valuation states diverge, people subsequently make different financial choices, even when the objective facts stay the same.

Standard financial models assume people experience time the same way regardless of what they're deciding. This research shows the brain actually stretches or compresses financial time based on valuation activity, which could explain why investors make inconsistent choices between mathematically equivalent options and why the same delay feels different in different emotional states.

Towards Computational Provenance: Carrying Causal-State Evidence in Generated Text

Hiding proof of how an AI reached its answer inside the answer itself

Researchers embedded hidden fingerprints into text generated by AI models that reveal which internal computational path the model used to reach its answer—even when different paths produce identical outputs. In controlled tests with both simple neural networks and transformers, a detector could later read these fingerprints and identify the verified internal state that was actually used, succeeding on all 128 test cases.

Today's AI systems are black boxes: you see the answer but not how the model produced it. If AI systems could cryptographically prove which computational steps they took, it would enable auditing, safety verification, and accountability—critical for high-stakes applications like medical diagnosis or financial decisions where knowing the reasoning matters as much as the result.