DdU1o2ZvWi / logbook.json
bertfil's picture
Backfill veritas replication logbook
dc3d644 verified
Raw
History Blame Contribute Delete
2.96 kB
{
"schema_version": 1,
"title": "Repro - Are VLMs Seeing or Just Saying? Uncovering the Illusion of Visual Re-examination",
"emoji": "\ud83d\udd2c",
"space_id": "bertfil/DdU1o2ZvWi",
"paper": {
"arxiv_id": "2605.15864",
"openreview_id": "DdU1o2ZvWi",
"title": "Are VLMs Seeing or Just Saying? Uncovering the Illusion of Visual Re-examination",
"url": "https://arxiv.org/abs/2605.15864"
},
"tags": [
"icml2026-repro",
"paper-DdU1o2ZvWi"
],
"updated_at": "2026-07-20T03:52:31+00:00",
"root": {
"slug": "index",
"title": "Repro - Are VLMs Seeing or Just Saying? Uncovering the Illusion of Visual Re-examination",
"file": "pages/index.md",
"children": [
{
"slug": "00-scored-evidence-summary",
"title": "00 - Scored evidence summary",
"file": "pages/00-scored-evidence-summary/page.md",
"children": []
},
{
"slug": "claim-1-under-the-visualswap-probing-protocol-qwen3-vl-235",
"title": "Claim 1 - Under the VisualSwap probing protocol, Qwen3-VL-235B-Thinkin\u2026",
"file": "pages/claim-1-under-the-visualswap-probing-protocol-qwen3-vl-235/page.md",
"children": []
},
{
"slug": "claim-2-averaged-across-models-visualswap-induces-an-accur",
"title": "Claim 2 - Averaged across models, VisualSwap induces an accuracy degra\u2026",
"file": "pages/claim-2-averaged-across-models-visualswap-induces-an-accur/page.md",
"children": []
},
{
"slug": "claim-3-vs-bench-comprises-800-curated-image-pairs-from-fo",
"title": "Claim 3 - VS-Bench comprises 800 curated image pairs from four benchma\u2026",
"file": "pages/claim-3-vs-bench-comprises-800-curated-image-pairs-from-fo/page.md",
"children": []
},
{
"slug": "claim-4-self-reflection-alone-produces-only-a-minimal-incr",
"title": "Claim 4 - Self-reflection alone produces only a minimal increase in vi\u2026",
"file": "pages/claim-4-self-reflection-alone-produces-only-a-minimal-incr/page.md",
"children": []
},
{
"slug": "claim-5-a-human-baseline-achieves-100-success-in-distingui",
"title": "Claim 5 - A human baseline achieves 100% success in distinguishing the\u2026",
"file": "pages/claim-5-a-human-baseline-achieves-100-success-in-distingui/page.md",
"children": []
},
{
"slug": "reproduction-protocol-and-provenance",
"title": "Reproduction protocol and provenance",
"file": "pages/reproduction-protocol-and-provenance/page.md",
"children": []
},
{
"slug": "conclusion",
"title": "Conclusion",
"file": "pages/conclusion/page.md",
"children": []
}
]
},
"agent_view_tokens": 8430,
"revision": "1784519551616104400"
}