We describe the Bubus submission to the SHROOM-Visions shared task on character-level hallucination span detection in image-conditioned LVLM outputs (English track). Our system fuses two complementary signals: (i) a cross-model supervised probe ensemble — per-layer logistic regressions over the hidden states of Qwen3-VL-32B and InternVL3-38B, each teacher-forcing the response under evaluation — and (ii) a K=8 role-diverse multi-sample LLM judge whose per-character marking fraction matches the annotator-fraction shape of the gold signal. A fixed two-weight late fusion with an abstain gate reaches Spearman correlation 0.411 and character IoU 0.381 on a leakage-controlled held-out split, measured once, post-competition (majority baseline 0.249). On the official test set the system scores Cor 0.413 / IoU 0.379 / Cor_lbl 0.282 — ranks 14/12/16 of 28 teams, the last on the primary ranking metric — agreeing with the held-out measurement within 0.002 on Cor and IoU, consistent with a leakage-free protocol. Beyond the positive recipe, we contribute (a) an unusually broad set of controlled negative results — zero-shot white-box signals all fail in the shared-task setting, which we attribute to the mismatch between self-generated and teacher-forced foreign text; (b) evidence that error-map-driven prompt engineering recovers precisely the targeted deficits at no precision cost; and (c) two external-data attempts that a pre-specified bootstrap gate correctly rejected.

We present BVAHAHA, a humor generationsystem for SemEval-2026 Task 1 (MWAHAHASubtask A), which frames constrained joke generation through the lens of Benign ViolationTheory (BVT). Given either two rare words ora news headline, the system generates contextually appropriate jokes while avoiding memorization and unsafe outputs. Our approachcombines BVT-guided humor generation witha parallel moderation pipeline (“Gatekeepers”)that detects excessive emotional intensity andhate speech, triggering iterative revisions whennecessary. Finally, we employ an LLM-as-aJudge framework with persona-based rankingto approximate human humor preferences.