zhinit's submissions | Hacker News

1.		Arbitrage Exists Between Kalshi and Polymarket but You Can't Exploit It (zhinit.dev)
		4 points by zhinit 46 days ago \| past
2.		YInMn Blue: The first new blue in two centuries (oregonstate.edu)
		13 points by zhinit 49 days ago \| past \| 2 comments
3.		Tool Specifications Matter: Uncovering and Mitigating Safety Risks in AI Agents (arxiv.org)
		2 points by zhinit 50 days ago \| past
4.		Stealthy Concurrent Audio Prompt Injections Against Multimodal LLM Agents (arxiv.org)
		2 points by zhinit 53 days ago \| past
5.		Are Kalshi MLB Game Winner Prices Accurate? (zhinit.dev)
		4 points by zhinit 55 days ago \| past
6.		Agentic Permissions Policy Algebra for Taint Confinement in LLM Agents (arxiv.org)
		2 points by zhinit 56 days ago \| past
7.		Convolution: The Secret Behind Filtering (thewolfsound.com)
		1 point by zhinit 61 days ago \| past
8.		Time-Frequency Consistency Learning for Robust Speech Deepfake Detection (arxiv.org)
		4 points by zhinit 63 days ago \| past
9.		How to Train a Gen AI Kick Drum Model on Your Old Linux Desktop with 6GB VRAM (zhinit.dev)
		165 points by zhinit 68 days ago \| past \| 86 comments
10.		Detecting Prompt Injection Attacks on Purpose-Specific LLM Agents (arxiv.org)
		3 points by zhinit 69 days ago \| past
11.		Prismata: Confining cross-site prompt injection in web agents (arxiv.org)
		14 points by zhinit 73 days ago \| past