{
  "id": 5657,
  "url": "https://arxiv.org/abs/2604.17475v1",
  "title": "Waking Up Blind: Cold-Start Optimization of Supervision-Free Agentic Trajectories for Grounded Visual Perception",
  "summary": "Small Vision-Language Models (SVLMs) are efficient task controllers but often suffer from visual brittleness and poor tool orchestration. They typically require expensive supervised trajectory tuning to mitigate these deficits. In this work, we propose Self-supervised Perception Enabled by Cascaded Tool Rollout Alignment (SPECTRA), a supervision-free framework that bootstraps agentic capabilities via Coldstart Reinforcement Learning for SVLMs. SPECTRA enforces Soft Structured Multi-turn Rollouts",
  "authors": "Ashutosh Bajpai, Tamal Majumder, Akshay Nambi, Tanmoy Chakraborty",
  "category": "research",
  "topics": "safety-alignment,agents-autonomy",
  "orgs": null,
  "regions": null,
  "published_at": "2026-04-19T15:06:30.000Z",
  "fetched_at": "2026-07-14T16:31:53.169Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/5657",
  "original_url": "https://arxiv.org/abs/2604.17475v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}