Jun 2026· arXiv.org· Vol abs/2606.26094· 0 citations· 53 references
Computer Science
TL;DR
RevengeBench, a benchmark of 75 LLM generated, Elo-calibrated policies across five game environments, drawn from CodeClash tournament trajectories, positions behavioral recovery of programmatic policies as a tractable inverse problem in code-space, opening a path to opponent modeling, policy interpretability, and the broader question of inferring latent mechanisms from observations.
Abstract
For most of scientific history, researchers studying behavior could only infer hidden mechanisms from outward actions: an inverse problem that becomes more tractable when observation is augmented by targeted intervention. We pose a computational analogue: given only behavioral traces of an agent in a game environment, can a learner reconstruct the underlying decision program as executable code, and how much does this reconstruction improve with the ability to design controlled experiments? We introduce RevengeBench, a benchmark of 75 LLM generated, Elo-calibrated policies across five game environments, drawn from CodeClash tournament trajectories. The learner observes the hidden target policy play against sampled opponents and designs behavioral probes in the form of custom opponent policies that elicit informative behavior. It then submits an executable hypothesis, which is evaluated using continuous action-distance metrics. We further validate that recovered code carries informative signal in downstream player-versus-player tournaments. Across twelve frontier LLMs, recovery quality varies substantially (34 to 72% of initial distance closed), with reconstructed policies yielding measurable competitive advantage, particularly for weaker models that otherwise struggle to design effective counter-strategies. Our benchmark positions behavioral recovery of programmatic policies as a tractable inverse problem in code-space, opening a path to opponent modeling, policy interpretability, and the broader question of inferring latent mechanisms from observations.
This work introduces a random variant sampler that applies common semantics-preserving transformations (SPTs) - spanning control-flow rewrites, dead-code injection, and identifier renaming - to produce perturbed variants, demonstrating that even top frontier models are susceptible to semantics-preserving perturbations.
Hasan Mahmud, Shreya Gupta, Isha Chaudhary et al.· 1 citation
Autonomous R\&D agents now write, run, and improve executable artifacts under automated evaluation---but largely as laboratory instruments: shown on curated benchmarks, with gains that are hard to trace to a cause and costs well above what sustained engineering practice absorbs. The limitation is structural. Most systems treat each attempt as nearly self-contained, so logs, memories, and search trees record what happened without establishing which design element produced an improvement, whether its evidence survived validation, or how it recombines with others. Long campaigns therefore keep re-learning the same lessons. We introduce Praxist, a lineage-centered generational system that converts reproducible artifacts and evaluator outcomes into a typed evidence graph of findings, lane-structured frontiers, and agendas. Separating local artifact construction from cohort-level evidence synthesis lets later attempts inherit validated mechanisms, unresolved claims, and useful constraints, and leaves results attached to an inspectable lineage. On the standardized 75-task MLE-bench suite, the finalized official-grader results give Praxist 60 medals (80.0\%), 49 of them gold, against 55 medals (73.3\%) and 34 gold for a Claude Code baseline on Claude Opus 4.8---at a recorded model spend of US\$3,054 versus US\$38,370, roughly a twelfth of the cost. Four case studies---quantitative trading, LiDAR-inertial-visual SLAM, tokamak magnetic control, and rocket landing---carry the same process into open-ended engineering problems, improving on each task-native baseline in headline accuracy, survival, or resource cost, with the discovery path on record. Stronger artifacts at an order of magnitude less spend, each backed by an auditable lineage, are, to our knowledge, first brought together here: the operating profile production research requires, not the one a benchmark demonstration establishes.
Jin Li, Ahmed Murtadha, Zhiying Wang et al.· 0 citations
Coding agents are ranked almost entirely by resolve rate: whether their final patch passes the target tests. Yet two agents can reach the same outcome through very different processes, and a single pass/fail label says nothing about why a run failed or why an accepted run spent extra steps, time, or tokens. This process evidence lives in the trajectory, which records a run's searches, reads, edits, tool calls, validation, and reversions. However, raw traces are heterogeneous and hard to compare across runs. We present TraceProbe, a trajectory-diagnostic framework that recovers what resolve rate hides. TraceProbe normalizes each raw run into a canonical nine-type action taxonomy with deterministic effect labels, then applies two rule-based modules: Insight names single-trajectory anti-patterns adapted from established debugging practice (e.g., search loops, verification skips), while Converge aligns pairs of runs and classifies where their behavior diverges under controlled references. Applying TraceProbe to 2,500 trajectories from five production settings on SWE-Bench Verified, we find that (i) file choice is too coarse to separate success from failure, whereas function selection and completion behavior localize it; (ii) Insight anti-patterns act mainly as corpus-level difficulty clues, with search loops the most stable; and (iii) even resolved runs differ in how quickly they reach relevant code and how much failed work they incur. Trajectory structure thus adds auditable diagnostic context to outcomes by localizing inspection targets, suggesting failure hypotheses, and prioritizing runs for review.
Rui Shu, Chun Yong Chong, Xin Zhou et al.· 1 citation
Together, these results recast the evolved harness as a legible compensation layer, shaped jointly by the language's engineering demands and the model's behavioral gaps, rather than an opaque benchmark-tuned scaffold.
Siqi Yang, Qianlan Yang, Yu-Xiong Wang et al.· 2 citations
This work proposes using reinforcement learning to synthesize human-readable Python programs that replicate an LLM’s input–output behavior, providing behavioral rather than internally faithful explanations.
ABE-Ralph is introduced, a reference-anchored auditing framework that represents claims, protocols, required components, baselines, and metrics as structured experimental constraints, guides implementation through an 8-step workflow, and performs quantitative, qualitative, and code-level verification.
Le-Zhi Yu, Xiao-Gang Xu, Yuhong Zhou et al.· 0 citations