DeepStack: Deeply Stacking Visual Tokens is Surprisingly Simple and Effective for LMMs · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2406.04334/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTDeepStack: Deeply Stacking Visual Tokens is Surprisingly Simple and Effective for LMMs — mentioned by nlhttps://news.ycombinator.com/item?id=45354802urn:hackerpapers:mention:2406.04334:45354802Wed, 24 Sep 2025 00:45:39 GMT<p>&gt; Take the core technology and just optimize, optimize, optimize for 10x the cost/efficiency. As simple as that. Super impressive. This &quot;just&quot; is incorrect. The Qwen team invented things like DeepStack https://arxiv.org/abs/2406.04334 (Also I hate this &quot;The Chinese&quot; thing. Do we say &quot;The British&quot; if it came from a DeepMind team in the UK? Or what if there are Chinese born US citizens working in Paris for Mistral? Give credit to the Qwen team rather than a whole country. China has both great labs and mediocre labs, just like the rest of the world.)</p><p><a href="https://hackerpaper.kaipfister.de/p/2406.04334/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.CV