Constraint Tax in Open-Weight LLMs: An Empirical Study of Tool Calling Suppression Under Structured Output Constraints
Paper • 2606.25605 • Published •
2
Aamer Mihaysi
O96a
·
AI & ML interests
Ethical AI, NLP & Cognitive architectures
Recent Activity
commentedon a paper about 4 hours ago
Constraint Tax in Open-Weight LLMs: An Empirical Study of Tool Calling Suppression Under Structured Output Constraints commentedon a paper 1 day ago
Escaping the Self-Confirmation Trap: An Execute-Distill-Verify Paradigm for Agentic Experience Learning commentedon a paper 2 days ago
Deeper is Not Always Better: Mitigating the Alignment Tax via Confident Layer DecodingOrganizations
commented a
paper about 4 hours ago
commented a
paper 1 day ago
commented a
paper 2 days ago
updated a
Space about 1 month ago
published a
Space about 1 month ago
updated a
Space about 1 month ago
published a
Space about 1 month ago
updated a
Space about 1 month ago
published a
Space about 1 month ago
updated a
Space about 1 month ago
published a
Space about 1 month ago
updated a
Space about 1 month ago
published a
Space about 1 month ago
updated a
Space about 1 month ago
published 2
Spaces about 1 month ago
updated 3
Spaces about 1 month ago
published a
Space about 1 month ago