כתבה
arXiv cs.AI ·
מציאת רעיונות מוזיקליים: גילוי קונספטים דרך התאמה של SAE
From Isolated Feature to Orbits: Discovering Music Concepts via Multi-SAE Alignment
במאמר זה, נציג פרקטיקה חדשה לגילוי קונספטים מוזיקליים, כולל חברות כמו LangGraph.
תקציר מקורי באנגליתarXiv:2610.01864v1 Announce Type: cross Abstract: How can we understand what a music foundation model has learned \textit{internally}? Most interpretability approaches, such as probing and Sparse Autoencoders (SAEs), focus on identifying individual features with minimal structural assumptions. We argue that many concepts are better understood as \textit{structured relations} rather than isolated features. This is especially prominent in music, where tonal structures are organized in the space of pitch and time. For example, concepts such as chords or keys are naturally expressed as structured sets (e.g., the 12 transpositions of a chord or the diatonic system within a key), rather than isolated features. In this study, \textbf{we shift from feature identification to structure-based analysi
קרא במקור המקורי
arxiv.org
פתח כתבה מקורית