PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents

Recursive self-improvement requires agents to turn accumulated experience into better future behavior. Personal AI agents offer a concrete setting for studying this capability because they retain preferences, task histories, tool routines, and learned skills across sessions. Yet whether retained experience actually improves them over time has not been systematically tested. We introduce PAST-Bench, a benchmark designed to isolate this question. Each agent runs through ordered sequences of fresh-session tasks under matched conditions that turn retained experience on and off. It spans 26 scenarios and 204 episodes across memory, procedural reuse, information gathering, and update. We report both later-task gains and whether those gains follow the intended save, retrieve, and update pathway. Across seven base models and four agent frameworks, improvement is real but uneven across capabilities. Agents with the same headline gain can differ markedly in whether that gain is supported by evidence of the intended pathway. Guided by these findings, we develop Hermes+, which extends Hermes with five targeted interventions across stages of the agent loop. Hermes+ raises the average gain from retained experience and provides clearer pathway evidence, with its strongest improvement on tasks requiring outdated state to be replaced, although the effect remains capability- and model-dependent. Together, PAST-Bench and Hermes+ provide an evaluation and diagnostic foundation for studying how persistent agents can progress from retaining experience to systematically improving through it. Code: https://github.com/Gen-Verse/PAST-Bench

Q-Score: A Quantum-Native Scoring Function for Molecular Docking

Molecular docking predicts how a small molecule binds to a protein and is a key bottleneck in drug discovery. Classical scoring functions sum empirical pairwise contacts, blind to quantum-mechanical effects like orbital charge transfer that govern binding specificity. We introduce Q-Score, encoding GNN-predicted orbital donor-acceptor energies into a weighted graph and scoring binding by solving a maximum-weight vertex clique problem via Digitized-Counterdiabatic QAOA. Each interaction anchor maps to one qubit and compatibility constraints become edges. Across 11 protein targets, DC-QAOA recovers the exact optimum on 8 at 10 qubits. On 1000 AI-generated molecules, Q-Score is orthogonal to classical scoring with Spearman rho of 0.05, driven by orbital quality with rho of 0.90, and free of molecular-weight bias, enriching for strong orbital interactions at twice the random rate. DC-QAOA achieves a mean approximation ratio of 0.94 with 52 percent exact. Execution of 1000 circuits on IBM Eagle confirms 6-qubit solvability on NISQ hardware.

Learning to Trigger: Reinforcement Learning at the Large Hadron Collider

High-throughput scientific facilities such as the Large Hadron Collider depend on real-time event filtering (\textit{triggering}) under tight constraints on bandwidth, latency, and storage. In practice, trigger menus are largely static and hand-tuned and can become suboptimal as detector conditions, pileup, and background composition drift over time. We cast online threshold tuning as a sequential decision-making problem: a reinforcement learning agent ingests streaming summaries of recent rates and signal-sensitive features and updates trigger thresholds to maximize signal efficiency while tracking a target background rate within a tolerance band. We adapt Group-Filtered Policy Optimization (GFPO) to streaming control and introduce two variants (GFPO-F, GFPO-FR) that enforce background rate feasibility during training. On a benchmark that emulates realistic collider operation, we study two representative triggers: a total transverse energy ($H_{T}$) trigger sensitive to pileup variation, and an anomaly-detection (AD) trigger based on reconstruction loss for rare or non-standard signatures. On Monte Carlo streams, our agent increases the fraction of in-tolerance time intervals by 48\% ($H_T$) and 28\% (AD), with a cumulative gain of up to 2\% in signal efficiency on those in-tolerance intervals. Transferring from simulation to \emph{real} collision data (CMS Run 283408), the same agent, without fine-tuning, achieves a 56\% ($H_T$) and 28\% (AD) in-tolerance improvement over baselines, with further signal-efficiency gain on both triggers. To our knowledge, this is the \emph{first} demonstration of RL-based trigger control on real Large Hadron Collider collision data. Code is available at https://github.com/Zixind/GFPO\_LHC.

THEMIS: Towards Holistic Evaluation of MLLMs for Scientific Paper Fraud Forensics

We present THEMIS, a novel multi-task benchmark designed to comprehensively evaluate multimodal large language models (MLLMs) on visual fraud reasoning within real-world academic scenarios. Compared to existing benchmarks, THEMIS introduces three major advances. (1) Real-World Scenarios and Complexity: Our benchmark comprises over 4,000 questions spanning seven scenarios, derived from authentic retracted-paper cases and carefully curated multimodal synthetic data. With 60.47% complex-texture images, THEMIS bridges the critical gap between existing benchmarks and the complexity of real-world academic fraud. (2) Fraud-Type Diversity and Granularity: THEMIS systematically covers five challenging fraud types and introduces 16 fine-grained manipulation operations. On average, each sample undergoes multiple stacked manipulation operations, with the diversity and difficulty of these manipulations demanding a high level of visual fraud reasoning from the models. (3) Multi-Dimensional Capability Evaluation: We establish a mapping from fraud types to five core visual fraud reasoning capabilities, thereby enabling an evaluation that reveals the distinct strengths and specific weaknesses of different models across these core capabilities. Experiments on 16 leading MLLMs show that even the best-performing model, GPT-5, achieves an overall performance of only 56.15%, demonstrating that our benchmark presents a stringent test. We expect THEMIS to advance the development of MLLMs for complex, real-world fraud reasoning tasks.

Bi-Level Prompt Optimization for Multimodal LLM-as-a-Judge

Large language models (LLMs) have become widely adopted as automated judges for evaluating AI-generated content. Despite their success, aligning LLM-based evaluations with human judgments remains challenging. While supervised fine-tuning on human-labeled data can improve alignment, it is costly and inflexible, requiring new training for each task or dataset. Recent progress in auto prompt optimization (APO) offers a more efficient alternative by automatically improving the instructions that guide LLM judges. However, existing APO methods primarily target text-only evaluations and remain underexplored in multimodal settings. In this work, we study auto prompt optimization for multimodal LLM-as-a-judge, particularly for evaluating AI-generated images. We identify a key bottleneck: multimodal models can only process a limited number of visual examples due to context window constraints, which hinders effective trial-and-error prompt refinement. To overcome this, we propose BLPO, a bi-level prompt optimization framework that converts images into textual representations while preserving evaluation-relevant visual cues. Our bi-level optimization approach jointly refines the judge prompt and the I2T prompt to maintain fidelity under limited context budgets. Experiments on four datasets and three LLM judges demonstrate the effectiveness of our method.

Towards a Self-Driving Trigger at the LHC: Adaptive Response in Real Time

Real-time data filtering and selection -- or trigger -- systems at high-throughput scientific facilities such as the experiments at the Large Hadron Collider (LHC) must process extremely high-rate data streams under stringent bandwidth, latency, and storage constraints. Yet these systems are typically designed as static, hand-tuned menus of selection criteria grounded in prior knowledge and simulation. In this work, we further explore the concept of a self-driving trigger, an autonomous data-filtering framework that reallocates resources and adjusts thresholds dynamically in real-time to optimize signal efficiency, rate stability, and computational cost as instrumentation and environmental conditions evolve. We introduce a benchmark ecosystem to emulate realistic collider scenarios and demonstrate real-time optimization of a menu including canonical energy sum triggers as well as modern anomaly-detection algorithms that target non-standard event topologies using machine learning. Using simulated data streams and publicly available collision data from the Compact Muon Solenoid (CMS) experiment, we demonstrate the capability to dynamically and automatically optimize trigger performance under specific cost objectives without manual retuning. Our adaptive strategy shifts trigger design from static menus with heuristic tuning to intelligent, automated, data-driven control, unlocking greater flexibility and discovery potential in future high-energy physics analyses.

Learning to Rank for Active Learning via Multi-Task Bilevel Optimization

Active learning is a promising paradigm to reduce the labeling cost by strategically requesting labels to improve model performance. However, existing active learning methods often rely on expensive acquisition function to compute, extensive modeling retraining and multiple rounds of interaction with annotators. To address these limitations, we propose a novel approach for active learning, which aims to select batches of unlabeled instances through a learned surrogate model for data acquisition. A key challenge in this approach is developing an acquisition function that generalizes well, as the history of data, which forms part of the utility function's input, grows over time. Our novel algorithmic contribution is a bilevel multi-task bilevel optimization framework that predicts the relative utility -- measured by the validation accuracy -- of different training sets, and ensures the learned acquisition function generalizes effectively. For cases where validation accuracy is expensive to evaluate, we introduce efficient interpolation-based surrogate models to estimate the utility function, reducing the evaluation cost. We demonstrate the performance of our approach through extensive experiments on standard active classification benchmarks. By employing our learned utility function, we show significant improvements over traditional techniques, paving the way for more efficient and effective utility maximization in active learning applications.