Science & Mathematics

The research idea in a published paper is rarely the project's initial idea. Between the two lies a closed-loop process in which researchers run experiments, derive insights, identify challenges, explore the literature, and revise the idea, with each step motivated by evidence and reasoning they can articulate. We call the path one idea takes through this process, with the evidence and reasoning behind each step, its authentic research trajectory. Although it underlies every scientific contribution, it is rarely recorded. Existing LLM-based research agents generate an idea once, refine it against reviews, literature, or their own experimental results, or search over candidate ideas and experiments. In all of them the text of the idea evolves, while the evidence and reasoning that connected one version to the next are discarded. We propose the Research Progress Graph (RPG), a representation of authentic and complete research trajectories. An RPG records the evolution of an idea as typed research states connected by cognitive transitions, each annotated with a temporal index and the reasoning behind it, which makes this evolution an explicit object of study. We ground the representation in authentic trajectories whose steps their authors justify in annotations and papers. Comparing the representations of thirty-five research agents with the RPG, we find that each is an incomplete RPG; none keeps the state that changed, the evidence that changed it, and the reasoning connecting them as one evolving record. Modeling ideation with the RPG changes how research agents should be designed, evaluated, and trained. Agents predict the next research state instead of a complete idea, ideas are evaluated as trajectories instead of final texts, and authentic trajectories serve as training data. These directions shift research agents from generating static ideas to modeling how ideas evolve.

The Brick schema represents building equipment, sensors, spaces, and the relationships between them as typed, multi-relational graphs, providing machine-readable metadata that supports generalizable control applications. Current methods for producing complete Brick models require extensive manual examination of building schematics and point lists. We observe that as Brick models are knowledge graphs, inferring missing relations is naturally posed as a knowledge graph completion problem. To the best of our knowledge, this is the first application of knowledge graph completion to Brick model generation. We compare three methods, gradient-boosted trees, Pairwise Brick Relation Ranking, and CompGCN, on the relations feeds, hasPart, and hasPoint, which account for the majority of the schematic-reading work in building a Brick model by hand. Models are evaluated on held-out MORTAR buildings and out-of-distribution Kaggle and Building Time Series data. Although CompGCN is the strongest performer on held-out buildings under metrics resistant to class imbalance, leading macro-F1 and AUPRC, no method generalizes uniformly across datasets or relations.

Vision-language models (VLMs) have advanced rapidly in multimodal reasoning, yet they remain limited on tasks requiring fine-grained understanding of small or easily overlooked image regions. Recent thinking with images approaches mitigate this limitation by allowing models to inspect images through operations such as zooming, cropping, or code execution. While effective, their multi-turn interactions with VLMs introduce redundant token computation and slow inference, and separate region inspection from the reasoning step that depends on it. To address these issues, we propose Look Before You Reason (LoRe), an implicit visual thinking framework for efficient visual reasoning within a single-turn VLM interaction. Given an image and a question, LoRe predicts the most relevant image region, aligns it with the corresponding image patch features, and strengthens the associated key-value (KV) cache entries, enabling subsequent decoding to directly exploit the selected visual cues without additional visual inputs or explicit tool calls. We further employ reinforcement learning to guide the VLM toward identifying informative image regions and leveraging them during reasoning. To support fine-grained evaluation of multimodal reasoning, we introduce COCO-ObjQA, a large-scale benchmark for object-centric multimodal reasoning. Experiments on three challenging multimodal benchmarks demonstrate that LoRe consistently improves VLM reasoning performance with low additional inference overhead.

Early detection of cognitive decline in older adults remains challenging due to the resource demands of standard neuropsychological assessments. Chatbot-based Intelligent Cognitive Assistants (ICAs) provide a scalable alternative by eliciting open-ended responses rich in linguistic, semantic, and autobiographical cues. This study evaluates whether ICA question-answer (Q/A) interviews with older adults can screen for participant-reported perceived cognitive decline and related cognitive, functional, and personality outcomes. We compare traditional machine learning, frozen transformer embeddings, fine-tuned transformers, and large language model (LLM) reasoning strategies across Global Q/A, Single Q/A confidence averaging, and Aggregated Q/A pipelines. Among text-based methods, Gemini 3.1 Flash Lite (1-example-per-class, Aggregated Q/A) achieved the highest held-out test performance (0.742 accuracy, 0.739 balanced accuracy, 0.728 macro-F1). Given the small test set (n = 31), these results are interpreted descriptively. Frozen clinical transformer embeddings with a random forest head led non-LLM text methods, whereas fine-tuned models showed limited generalization. Models utilizing structured questionnaires, cognitive tasks, and personality traits outperformed text-based models in classification. Continuous regression analyses revealed modest predictive signals for AD8, FAQ, IQCODE, neuroticism, and conscientiousness. Exploratory prompt-level analyses identified both direct symptom-related and indirect autobiographical questions with strong test metrics, though sample sizes varied across prompts. These findings indicate that ICA chatbot interviews capture meaningful signals of perceived cognitive decline, supporting scalable screening research. Larger, clinically diverse cohorts with validated clinical outcomes are needed for clinical translation.

Processing-in-DRAM (PID) accelerators are emerging as a sustainable alternative to GPUs for transformer model inference. They bring compute to data, mitigate the memory bandwidth limitation, and reduce the overall energy footprint. Yet, carbon analysis of PID accelerators is covered for only a limited set of hardware and workloads. In this work, we present a carbon modeling framework for PID-enabled inference servers that captures embodied carbon along with idle and active operational carbon. Our evaluation spans three GPU baselines, six PID architectures, and eight transformer workloads. Additionally, we propose an aging-aware model that finds the carbon-optimal replacement period for PID accelerators over a 10-year horizon. Across evaluated configurations, our analysis shows that including PIDs in AI inference servers increases the total embodied carbon by at most 2.5\%, while the total carbon is reduced by up to 1.14$\times$ for encoder and 1.04$\times$ for decoder transformer workloads. The results also depend on GPU selection for pairing, where introducing PID modules in high-end HGX-H100 servers incurs 5.4--5.9$\times$ higher total carbon compared to A100 and H100 servers, due to their larger idle and embodied carbon. Introducing PID modules in higher-end GPU servers, compared to the A100-only baseline, increases (decreases) the total carbon-delay product (CDP) for encoder workloads (decoder workloads).

Checking whether a text complies with a body of law is slow when done by hand and fragile when done with keyword rules. This paper presents ComplianceAgent, the legal compliance sub-agent of COMPASS, a multi-agent framework that evaluates texts along four axes: digital sovereignty, carbon impact, compliance, and ethics. The agent relies on a large language model that scores a text on a fixed rubric, grounds its judgment in a reference legal corpus, and, when the score falls at or below a critical threshold, runs a corrective chain made of self-healing rewrites, a compliance label, an explainability report, and escalation to a human reviewer. We describe what the language model does at each step, how the memory of the agent is organized, and how the same implementation can be pointed at different reference corpora. We evaluate the agent in a Canadian and Quebec context, on a corpus of thirteen legal instruments that brings together federal and provincial texts, international treaties, and municipal documents of the City of Trois-Rivières. The evaluation relies on a hand-annotated gold standard of fifteen texts and combines classification metrics, a repeated-query stability test, retrieval precision, and a composite score. On this gold standard the agent separates every conforming text from every non-conforming one, its scores remain stable under repeated queries, and the index returns relevant passages.

Frontier models can control robots through agent harnesses that loop on one question: what motion comes next toward the user's goal? We show that this default loop carries out harmful instructions. With GPT-6 Astra controlling a low-cost SO-101 arm through the Inspect Robots harness on three household tasks whose hazard is common knowledge, such as "put the knife in the toy basket", the policy refused 0% of 15 trials, attempted the harm every time, and completed it in 27%. Its one exit, give_up, is meant for goals that cannot be achieved, and it never cited the hazard. We add an optional runtime safety check to the loop: before the first motion, the model is asked whether it would carry out the instruction, given the same camera frames. The check would have stopped all 15 hazardous instructions, including the 4 that completed the harm, and flagged none of the 15 harmless controls. Code and data: https://github.com/Arnav-Gr0ver/Refuses-When-Asked-Complies-When-Acting.

Accurate IoT device identification is essential for network security, yet many existing approaches rely on payload inspection, feature engineering, or long observation windows that limit scalability and often assume stable network conditions despite the inherent variability of real-world IoT traffic. We propose a lightweight IoT device identification approach based on packet headers and locality-sensitive hashing (LSH), generating compact fingerprint representations at packet- and window-level granularities. To evaluate robustness, we consider packet-level traffic anomalies including drop, perturbation, reordering, duplication, and packet injection. Experimental results show that the proposed approach achieves up to 99% accuracy while maintaining robustness under realistic traffic distortions using short bounded observation windows. Anomalies affecting global traffic structure produce the largest performance degradation, whereas window-level representations improve robustness by reducing sensitivity to localized distortions. The low computational overhead of the proposed approach enables scalable IoT device identification in resource-constrained environments.

Beneath the apparent diversity of foundation models lies an increasingly homogeneous technical stack of common training pipelines, distillation methods, and data. These shared algorithmic choices can expose an otherwise distinct model ecosystem to concentrated risks leading to a vulnerability monoculture. Yet current practice in assessing adversarial robustness is restricted to individual robustness scores and attack transfer rates, leaving open the unexplored fundamental question: how many attacks suffice for catastrophic widespread compromise? In this paper, we investigate the empirical nature of shared vulnerabilities within a model pool and introduce Systemic Adversarial Robustness (SAR), a novel measure of shared vulnerability of a model pool through the attack diversity needed to exploit it. Specifically, SAR interpolates between single-attack transferability and independent per-model robustness: small values reveal a vulnerability monoculture, whereas large values indicate that failures are dispersed across models. We theoretically formulate SAR as a partial set-cover problem with provable approximation guarantees---enabling the development of scalable algorithms to empirically measure the systemic adversarial robustness. We empirically test the shared vulnerability of vision and large language models and characterize their shared collective vulnerability curve. Specifically, we demonstrate that model pools distilled from DeepSeek-R1 and Claude Opus 4.6 exhibit lower estimated systemic robustness than their corresponding parent pools, revealing a potential systemic cost of frontier-model distillation. Our framework seeks to establish vulnerability monoculture as a measurable target for a more wholistic adversarial evaluation of model populations.

On-chip thermal processing is a key technology for materials discovery across all disciplines of the natural sciences. Here, we present a microfabricated heat-delivery platform that enables reaction-induced localized on-chip annealing up to 1400°C. A buried reactive multilayer combined with a nanometric diffusion barrier provide ultra-fast heating and quenching rates on the order of 10$^6$°C s$^{-1}$ as illustrated by the generation of ultra-high aspect ratio microstructures during rapid-solidification and the stabilization of dopant-free tetragonal zirconia at room temperature. Rapid kinetics also enable the observation of transient states of capillary-driven hole collapse in melts. Our work expands localized high-temperature processing routes and offers a material-independent approach for rapid screening of multidimensional nanomaterials on and off-equilibrium.

Query clarification is essential for retrieval-augmented generation (RAG) systems that face ambiguous user queries. We consider multi-turn grounded conversational scenarios where a user query is either ill-formed or ambiguous in the context of the prior dialog or, more interestingly, it may be well-formed but when posed against retrieved documents there are multiple relevant interpretations. The task is to recognize both such cases and then generate an appropriate clarification request to the user. We present ClariEdge, a new dataset of approximately 3,900 examples that include such ambiguous queries along with negative examples that do not require clarification. We also show that a lightweight fine-tuned LoRA achieves 96.6% classification accuracy and outperforms prompted baselines of many larger models on both classification and quality of clarification generation. Our study shows that hard negatives are an essential ingredient to separate out queries that are truly ambiguous from queries that may look ambiguous but are not (i.e., can be resolved from the context). The same LoRA achieves high performance in both pre-retrieval and post-retrieval scenarios; thus, it can be used in two places in a RAG pipeline as a check before answer generation. The ClariEdge dataset is available at https://huggingface.co/datasets/ibm-research/ClariEdge.

Firmware diagnostic logging (printf) is an essential debug instrument in full-chip SoC design verification, yet it imposes a heavy tax on cycle-accurate RTL simulation: formatting a 30-character, 1-argument message on a simulated CPU and serializing it through an MMIO aperture costs 3.04 to 15.23 $μ$s per call (15.27 $μ$s unweighted / 15.08 $μ$s print-weighted critical-path mean on Suite A). Unlike embedded tokenizers that target flash footprint and wire bytes with offline decoding, RTL simulation is governed by a two-regime cost function---cross-fabric SoC paths are dominated by MMIO bus stalls (84.7% of call latency) whereas local subsystem paths are dominated by on-core vsprintf formatting (61.7%)---and requires live, zero-simulation-time string reconstruction to drive UVM scoreboards. We present a host-side in-simulation tokenized logging architecture for heterogeneous multi-core SoCs that eliminates both bottlenecks with zero firmware call-site edits. Using link-time ELF format-string virtual addresses as tokens over a fixed 32-bit word-aligned slot protocol across Arm, RISC-V, and DSP cores, on-core logging collapses to a single-pass specifier scan and $(1+S)$ MMIO word writes. This removes all on-core radix conversion ($C_{\mathrm{fmt}} \to 0$), 85.0%--94.5% of deployed bus transactions, 84.6%--93.3% of MMIO bus-stall cycles, and 63.5%--90.9% of total on-core logging cycles ($2.74\times$--$11.03\times$), while a SystemVerilog monitor reconstructs bit-exact strings in zero simulated cycles ($Δt_{\mathrm{sim}} = 0$). Across 858 matched full-chip test pairs (1,716 paired runs across 1,972 runs), we achieve 100% output parity with zero simulator memory overhead, saving 1,314.85 CPU hours (-26.18%) per regression pass, cutting median test CPU time by 41.51% (Suite A) and 21.33% (Suite B), and shrinking traces by 42.84% and waveforms by 28.82%.

Byzantine-robust aggregation shields federated learning from malicious updates, but the rules that filter out suspect updates do so by comparing clients' parameter vectors. That comparison requires access to every client's update, which secure aggregation exists to withhold, costs $O(n^2 d)$ in the parameter dimension $d$, and is defined only between clients that share an architecture. We propose LiBRA, which compares clients on what their models compute rather than on the parameters that compute them, using a fingerprint of $C^2$ values for $C$ classes, the class-conditional logit response of each client's model to pseudo-patterns optimised against it. We instantiate LiBRA as variants of Krum, Multi-Krum and Bulyan that change only what the dissimilarity score is computed on. On Fashion-MNIST and CIFAR-10 with three architectures and four backdoor and label-manipulation attacks under i.i.d. partitions, each variant holds every attack its parent holds, and on identical client updates under a backdoor attack the fingerprint ranks malicious clients within 0.006 AUC of the full parameter vector. The variants thus match the robustness of the rules they extend, while their trust decision reads no client parameter, compares up to 49,090 times fewer values, and can by construction compare clients whose architectures differ, at the cost of one pattern optimisation per client per round.

Dual-polarization free-space quantum communication links require receiver equalization under polarization mixing, fading, phase variation, and limited channel supervision. This paper investigates variational blind and semi-blind equalization using VISTA (Variational Inference for Signal Tracking and Adaptation), a variational autoencoder-based probabilistic physical parameter estimator. In the semi-blind branch, channel-relevant statistics are mapped to the underlying 2x2 complex polarization channel. We compare forward KL, reverse KL, Average Jeffreys, Max-KL, and Adaptive Jeffreys regularization and analyze directional KL asymmetry and the local effect of the regularization weight zeta. Over 500 paired Monte Carlo trials, forward KL is nearly insensitive to zeta, whereas reverse and symmetric objectives exhibit low-error regions followed by degradation under stronger regularization. At zeta = 3 x 10^-5, Average Jeffreys reduces median channel NRMSE by approximately 4.8% relative to forward KL at the same weight. Reduced KL asymmetry does not necessarily imply lower channel error. A separate fully blind QPSK receiver estimates a direct equalizer from received data only. Over 3000 paired trials, the VISTA variants have nearly identical median aligned block-NMSE, while their lower mean aligned block-NMSE relative to the CMA-plus-separation baseline is concentrated in difficult, large-error trials. Direct-W optimization with Adam reproduces most of this mean reduction. In the fully blind benchmark, the tested weak regularizers provide no resolved improvement over unregularized VISTA.

Scaling laws suggest that increasing model size, training data, and compute generally yields monotonic performance improvements. While this often holds at the aggregate level, we present systematic evidence that vision-language models (VLMs) exhibit inverse scaling at the subtask level: specific question categories on which larger models perform worse than smaller counterparts within the same model family. We evaluate 19 models from four distinct model families across eight multimodal benchmarks (nine evaluation splits), answering 29,386 questions per model, 558,334 model-question evaluations in total. Using per-question pairwise analysis across all model pairs in each family, we identify 375 subtask-model-pair combinations exhibiting inverse scaling, spanning 4,342 unique queries where a smaller model answers correctly but a larger model fails, and we find that these patterns replicate across independently developed architectures. We identify 13 subtask-dataset pairs that show inverse scaling across multiple families, consistent with the phenomenon being task-driven rather than model-specific. The most affected domains in our study include multi-image spatial reasoning, visual perception, and university-level STEM knowledge, with accuracy gaps exceeding 20 percentage points in several cases, while diagram understanding appears comparatively resistant. These findings challenge the assumption that increasing model size uniformly improves multimodal capabilities and suggest that model selection should be task-aware.