Saved in:
Bibliographic Details
Main Author: Stasiuc, Victor
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.17956
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908736509968384
author Stasiuc, Victor
author_facet Stasiuc, Victor
contents Safety alignment can make frontier LMs overly conservative, degrading collaboration via hedging or false refusals. We present a lightweight toolkit with three parts: (1) Victor Calibration (VC), a multi-pass protocol that elicits a scalar confidence proxy T (T0<T1<T2) through iterative evidence re-evaluation; (2) FD-Lite, a behavior-only phenomenology audit with a fixed anchor phrase and a meta-prefix trap to avoid anthropomorphic claims; and (3) CP4.3, a governance stress test for rank invariance and allocation monotonicity (M6). Across Claude 4.5 models (Haiku, Sonnet no-thinking, Sonnet thinking) and Opus, we observe monotonic VC trajectories without violating safety invariants, and stable CP4.3 behavior. ("Opus" here refers to a single Claude Opus 4.1 session accessed via a standard UI account, as reported in Table 1.) This work was conducted by a single operator (n=1) and is intended as hypothesis-generating; we explicitly invite replication, critique, and extension by the research community. We include prompt templates and an artifact plan to facilitate independent verification.
format Preprint
id arxiv_https___arxiv_org_abs_2512_17956
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Victor Calibration (VC): Multi-Pass Confidence Calibration and CP4.3 Governance Stress Test under Round-Table Orchestration
Stasiuc, Victor
Software Engineering
Artificial Intelligence
Safety alignment can make frontier LMs overly conservative, degrading collaboration via hedging or false refusals. We present a lightweight toolkit with three parts: (1) Victor Calibration (VC), a multi-pass protocol that elicits a scalar confidence proxy T (T0<T1<T2) through iterative evidence re-evaluation; (2) FD-Lite, a behavior-only phenomenology audit with a fixed anchor phrase and a meta-prefix trap to avoid anthropomorphic claims; and (3) CP4.3, a governance stress test for rank invariance and allocation monotonicity (M6). Across Claude 4.5 models (Haiku, Sonnet no-thinking, Sonnet thinking) and Opus, we observe monotonic VC trajectories without violating safety invariants, and stable CP4.3 behavior. ("Opus" here refers to a single Claude Opus 4.1 session accessed via a standard UI account, as reported in Table 1.) This work was conducted by a single operator (n=1) and is intended as hypothesis-generating; we explicitly invite replication, critique, and extension by the research community. We include prompt templates and an artifact plan to facilitate independent verification.
title Victor Calibration (VC): Multi-Pass Confidence Calibration and CP4.3 Governance Stress Test under Round-Table Orchestration
topic Software Engineering
Artificial Intelligence
url https://arxiv.org/abs/2512.17956