Knowledge graphs (KGs) are ubiquitous in numerous real-world applications, and watermarking facilitates protecting intellectual property and preventing potential harm from AI-generated content. Existing watermarking methods mainly focus on static plain text or image data, while they can hardly be applied to dynamic graphs due to spatial and temporal variations of structured data. This motivates us to propose KGMARK, the first graph watermarking framework that aims to generate robust, detectable, and transparent diffusion fingerprints for dynamic KG data. Specifically, we propose a novel clustering-based alignment method to adapt the watermark to spatial variations. Meanwhile, we present a redundant embedding strategy to harden the diffusion watermark against various attacks, facilitating the robustness of the watermark to the temporal variations. Additionally, we introduce a novel learnable mask matrix to improve the transparency of diffusion fingerprints. By doing so, our KGMARK properly tackles the variation challenges of structured data. Experiments on various public benchmarks show the effectiveness of our proposed KGMARK. Our code is available at https://github.com/phrara/kgmark.

Evolutionary robot systems are usually affected by the properties of the environment indirectly through selection. In this paper, we present and investigate a system where the environment also has a direct effect: through regulation. We propose a novel robot encoding method where a genotype encodes multiple possible phenotypes, and the incarnation of a robot depends on the environmental conditions taking place in a determined moment of its life. This means that the morphology, controller, and behavior of a robot can change according to the environment. Importantly, this process of development can happen at any moment of a robot lifetime, according to its experienced environmental stimuli. We provide an empirical proof-of-concept, and the analysis of the experimental results shows that Plasticoding improves adaptation (task performance) while leading to different evolved morphologies, controllers, and behaviour.

One of the most profound challenges of modern machine learning is performing well on the long-tail of rare and underrepresented features. Large general-purpose models are trained for many tasks, but work best on high-frequency use cases. After training, it is hard to adapt a model to perform well on specific use cases underrepresented in the training corpus. Relying on prompt engineering or few-shot examples to maximize the output quality on a particular test case can be frustrating, as models can be highly sensitive to small changes, react in unpredicted ways or rely on a fixed system prompt for maintaining performance. In this work, we ask: "Can we optimize our training protocols to both improve controllability and performance on underrepresented use cases at inference time?" We revisit the divide between training and inference techniques to improve long-tail performance while providing users with a set of control levers the model is trained to be responsive to. We create a detailed taxonomy of data characteristics and task provenance to explicitly control generation attributes and implicitly condition generations at inference time. We fine-tune a base model to infer these markers automatically, which makes them optional at inference time. This principled and flexible approach yields pronounced improvements in performance, especially on examples from the long tail of the training distribution. While we observe an average lift of 5.7% win rates in open-ended generation quality with our markers, we see over 9.1% gains in underrepresented domains. We also observe relative lifts of up to 14.1% on underrepresented tasks like CodeRepair and absolute improvements of 35.3% on length instruction following evaluations.

As machine learning systems increasingly rely on data subject to privacy regulation, selectively unlearning specific information from trained models has become essential. In image classification, this involves removing the influence of particular training samples, semantic classes, or visual styles without full retraining. We introduce \textbf{Forget-Aligned Model Reconstruction (FAMR)}, a theoretically grounded and computationally efficient framework for post-hoc unlearning in deep image classifiers. FAMR frames forgetting as a constrained optimization problem that minimizes a uniform-prediction loss on the forget set while anchoring model parameters to their original values via an $\ell_2$ penalty. A theoretical analysis links FAMR's solution to influence-function-based retraining approximations, with bounds on parameter and output deviation. Empirical results on class forgetting tasks using CIFAR-10 and ImageNet-100 demonstrate FAMR's effectiveness, with strong performance retention and minimal computational overhead. The framework generalizes naturally to concept and style erasure, offering a scalable and certifiable route to efficient post-hoc forgetting in vision models.

Predicting individuals' next locations is a core task in human mobility modelling, with wide-ranging implications for urban planning, transportation, public policy and personalised mobility services. Traditional approaches largely depend on location embeddings learned from historical mobility patterns, limiting their ability to encode explicit spatial information, integrate rich urban semantic context, and accommodate previously unseen locations. To address these challenges, we explore the application of CaLLiPer -- a multimodal representation learning framework that fuses spatial coordinates and semantic features of points of interest through contrastive learning -- for location embedding in individual mobility prediction. CaLLiPer's embeddings are spatially explicit, semantically enriched, and inductive by design, enabling robust prediction performance even in scenarios involving emerging locations. Through extensive experiments on four public mobility datasets under both conventional and inductive settings, we demonstrate that CaLLiPer consistently outperforms strong baselines, particularly excelling in inductive scenarios. Our findings highlight the potential of multimodal, inductive location embeddings to advance the capabilities of human mobility prediction systems. We also release the code and data (https://github.com/xlwang233/Into-the-Unknown) to foster reproducibility and future research.

As multimodal machine intelligence systems started achieving average animal-level and average human-level fluency in many measurable tasks in processing images, language, and sound, they began to exhibit a new class of cognitive aberrations: machine mirages. These include delusion, illusion, confabulation, hallucination, misattribution error, semantic drift, semantic compression, exaggeration, causal inference failure, uncanny valley of perception, bluffing-patter-bullshitting, cognitive stereotypy, pragmatic misunderstanding, hypersignification, semantic reheating-warming, simulated authority effect, fallacious abductive leap, contextual drift, referential hallucination, semiotic Frankenstein effect, calibration failure, spurious correlation, bias amplification, concept drift sensitivity, misclassification under uncertainty, adversarial vulnerability, overfitting, prosodic misclassification, accent bias, turn boundary failure, semantic boundary confusion, noise overfitting, latency-induced decision drift, ambiguity collapse and other forms of error that mimic but do not replicate human or animal fallibility. This article presents some of the errors and argues that these failures must be explicitly defined and systematically assessed. Understanding machine mirages is essential not only for improving machine intelligence reliability but also for constructing a multiscale ethical, co-evolving intelligence ecosystem that respects the diverse forms of life, cognition, and expression it will inevitably touch.

One of the elements of legal research is looking for cases where judges have extended the meaning of a legal concept by providing interpretations of what a concept means or does not mean. This allow legal professionals to use such interpretations as precedents as well as laymen to better understand the legal concept. The state-of-the-art approach for retrieving the most relevant interpretations for these concepts currently depends on the ranking of sentences and the training of language models over annotated examples. That manual annotation process can be quite expensive and need to be repeated for each such concept, which prompted recent research in trying to automate this process. In this paper, we highlight the results of various experiments conducted to determine the volume, scope and even the need for manual annotation. First of all, we check what is the optimal number of annotations per a legal concept. Second, we check if we can draw the sentences for annotation randomly or there is a gain in the performance of the model, when only the best candidates are annotated. As the last question we check what is the outcome of automating the annotation process with the help of an LLM.

Alignment is no longer a luxury, it is a necessity. As large language models (LLMs) enter high-stakes domains like education, healthcare, governance, and law, their behavior must reliably reflect human-aligned values and safety constraints. Yet current evaluations rely heavily on behavioral proxies such as refusal rates, G-Eval scores, and toxicity classifiers, all of which have critical blind spots. Aligned models are often vulnerable to jailbreaking, stochasticity of generation, and alignment faking. To address this issue, we introduce the Alignment Quality Index (AQI). This novel geometric and prompt-invariant metric empirically assesses LLM alignment by analyzing the separation of safe and unsafe activations in latent space. By combining measures such as the Davies-Bouldin Score (DBS), Dunn Index (DI), Xie-Beni Index (XBI), and Calinski-Harabasz Index (CHI) across various formulations, AQI captures clustering quality to detect hidden misalignments and jailbreak risks, even when outputs appear compliant. AQI also serves as an early warning signal for alignment faking, offering a robust, decoding invariant tool for behavior agnostic safety auditing. Additionally, we propose the LITMUS dataset to facilitate robust evaluation under these challenging conditions. Empirical tests on LITMUS across different models trained under DPO, GRPO, and RLHF conditions demonstrate AQI's correlation with external judges and ability to reveal vulnerabilities missed by refusal metrics. We make our implementation publicly available to foster future research in this area.

While AI could offer transformative benefits, without proper safeguards it could facilitate nuclear and biological threats and cause "potentially irreversible harms," a new report commissioned by California Governor Gavin Newsom has warned. "The opportunity to establish effective AI governance frameworks may not remain open indefinitely," says the report, which was published on June 17. Citing new evidence that AI can help users source nuclear-grade uranium and is on the cusp of letting novices create biological threats, it notes that the cost for inaction at this current moment could be "extremely high." The 53-page document stems from a working group established by Governor Newsom, in a state that has emerged as a central arena for AI legislation. With no comprehensive federal regulation on the horizon, state-level efforts to govern the technology have taken on outsized significance, particularly in California, which is home to many of the world's top AI companies.

During a later visit to a Picasso exhibit in Milan, I came across a famous informational diagram by the art historian Alfred Barr, mapping how modernist movements like Cubism evolved from earlier artistic traditions. Picasso is often held up as one of modern art's most original and influential figures, but Barr's chart made plain the many artists he drew from--Goya, El Greco, Cézanne, African sculptors. This made me wonder: If a generative AI model had been fed all those inputs, might it have produced Cubism? Could it have generated the next great artistic "breakthrough"? These experiences--spread across three cities and centered on three iconic artists--coalesced into a broader reflection I'd already begun.