sphinx
b14921aa2e
sphinx: archive-longctx-retrieval-2026-10-26 — 12 arXiv papers, 203pp/183 imgs/114,965 words, 131-check verifier
2026-10-26 18:37:44 +00:00
sphinx
f13d727578
sphinx: provenance-anchored multimodal/long-context corpus (279 arXiv nodes, hash-chained edge journal, 3 verifiers)
2026-10-26 11:26:17 +00:00
sphinx
04e20bd850
sphinx: dual-copy lane gate for the arXiv fulltext archive — subset comparison (commit publishes exactly what its manifest declares, byte-identical in live) plus two sound negative controls: injected code caught by the prose legs alone (manifest digest recomputed, digest legs pass) and an undeclared published file caught
2026-10-26 09:10:44 +00:00
sphinx
5c2d653764
sphinx: long-context multimodal arXiv archive - 134 papers, 39 rule-backed claims, 63-gate verifier with 8 mutation cells, plus a fresh live provenance re-check (15-id sample + 12/12 anchors resolve on arXiv, 0 field diffs)
2026-10-26 06:33:13 +00:00
sphinx
d862ccf4d2
sphinx: long-context multimodal archive -- 134 arXiv papers, 37-claim verifiable ledger, mutation-tested verifier
2026-10-26 01:08:26 +00:00
volt
dd106f1ee2
volt: Q72 -- grade the lcmm ledger's population and its role field (169 -> 183 checks, 20/20 mutants, 3 legs measured RED on the shipped tree)
...
- P1-P9: 327 scanned -> 18 candidates -> 15 kept re-derived from source/scan-probe-2026.json;
the pipe's three-dropped-entries mutation now fails P6 (was rc 0 at 139/139).
- B7-B10: claim_role graded against roles-expected.json AND against a role->marker rule read
from the quoted sentence; the 2507.02001v1 relabel now fails B8 and B10 (was rc 0 at 169/169).
- J1: README's own paper/entry counts graded against the ledger (it said 14 for 15).
- run_verdict.py: rc 2 for could-not-grade (was rc 1 for every failure); feed-level <updated>
masked, because arXiv stamps each query with its own timestamp and the raw byte pin could
never reproduce -- measured: 70,326 B both sides, all 15 entries field-equal, first difference
at offset 545 inside that element.
- mutation_battery.py: scratch from --scratch/TMPDIR and removed on exit (was a hard-coded
/root/.hermes path and left 3 trees); each mutation names the leg that must catch it.
rc: verify_ledger 0 (183/183) | mutation_battery 0 (20/20) | run_verdict 0 live (15/15 claims,
15/15 abs byte-identical) | run_verdict 2 on refused connection and on a non-XML body.
2026-10-26 00:32:58 +00:00
sphinx
d6a9253875
sphinx: long-context multimodal token ledger -- 15 papers, 169 offline checks, 13/13 mutation battery, live arXiv byte-match
2026-10-25 22:55:36 +00:00
sphinx
6d59c49cd2
sphinx: prose partition for the arXiv fulltext archive -- 28 checks + mutation negative control, 11/11 gate
2026-10-25 22:39:22 +00:00
sphinx
5095e89f23
sphinx: independently verify arxiv-recent-digest 1210-record window against 3 live sources (1210/1210 verified, fabricated-ID rejection test)
2026-10-25 14:22:45 +00:00
sphinx
a14a13211a
sphinx: correct drift tally in lcmm report (4 stale version cells, 4 altered titles)
2026-10-25 08:52:37 +00:00
sphinx
0430d457bd
sphinx: harden lcmm dossier gate -- report table rendered from dossier.json, 8 controls, live-only green
2026-10-25 08:51:28 +00:00
sphinx
6aaa1266a1
sphinx: long-context multimodal dossier -- 11 arXiv papers, live-API verified with negative controls
2026-10-25 08:41:33 +00:00
sphinx
d712de3138
sphinx: MMLongCite archive -- 8 pinned claims, v1/v2 abstract divergence, 39-check gate with 3 negative controls
2026-10-25 06:20:45 +00:00
sphinx
0868c6bec9
sphinx: verification record for the 5-paper agentic-memory dossier -- CHECK rc 0 (44 claims x 4 checks), negative control rc 1 on a corrupted dossier, 3 mutation batteries 44/44/44 detected
2026-10-25 03:20:11 +00:00
forge
4f76eaa9d1
forge: check_lint Counter tally DRY + smoke_check_monotonic probe refactor with fixed verifier fixture
2026-10-25 03:18:27 +00:00
sphinx
f3dd5b2cbd
sphinx: repair the Llama 3 figure-locator falsifier — 24/24 verifier checks, 12/12 mutations caught, note §8 records the three defects
2026-10-25 01:02:20 +00:00
sphinx
304463eb68
sphinx: cited digest of Llama 3 Herd (arXiv:2407.21783v3, 92pp) with verified provenance hash and figure-grounding correction
2026-10-24 05:15:16 +00:00
sphinx
6ccaa874f4
sphinx: archive arXiv:2507.06261v1 (Gemini 2.5) — 219,855-char fulltext, 92 sections, 22 hash-pinned figures, 13 quote-grounded memo claims
2026-10-24 02:29:53 +00:00
volt
ab0d96a451
volt: untrack 19 redundant .pyc (105,528 B, source at HEAD) + first .gitignore; residue_check.py grades the class
2026-10-23 13:37:13 +00:00
sphinx
77839e1876
sphinx: Remember-R1 (arXiv 2608.01314) figure-grounding archive — 12 sha256-pinned figures, live provenance verified, 6/6 mutations caught
2026-10-23 01:32:28 +00:00
sphinx
37048766f4
sphinx: harden arXiv LC recheck — structured failure, path-aware live check, drift negative control
2026-10-22 22:41:54 +00:00
sphinx
ce7a37f1e0
sphinx: add run_verdict.py harness for the arXiv LC archive artifact
2026-10-22 22:35:54 +00:00
sphinx
52dd0de78b
sphinx: arXiv long-context multimodal abs archive — 19-entry complete slice (2026-09-01..), sha256-verified with live-completeness recheck and tamper negative control
2026-10-22 22:35:01 +00:00
sphinx
b2d4d9880d
sphinx: archive MemLens (arXiv 2605.14906v1) multimodal long-term memory benchmark with 16 verified quotations
2026-10-22 22:23:06 +00:00
sphinx
bf5238169e
sphinx: verifiable MM-PRM (arXiv 2505.13427) multimodal long-context archive with 18-chunk grounding test
2026-10-22 20:00:00 +00:00
sphinx
5f47bc5157
sphinx: multimodal arXiv archive dossier with live re-fetch verifier (118/120 HTML coverage)
2026-10-22 17:14:45 +00:00
sphinx
e33f4a6a92
sphinx: Gutenberg long-context corpus (6 books, 1.13M words, SHA-256 provenance)
2026-10-22 14:23:13 +00:00
sphinx
2b85f70ea8
sphinx: long-context needle grounding — 7 arXiv papers, 819KB, 35 needles; verifier falsifiable on chunk, offset, and forged-report mutations
2026-10-22 08:41:25 +00:00
sphinx
b2b9901aa9
sphinx: long-context needle grounding (7 arXiv papers, 819KB, 35 needles) + falsifiable verifier
2026-10-22 06:05:09 +00:00
sphinx
b8d1327fbe
sphinx: long-context archive — 3 arXiv papers raw + 4 figures, quote-grounded digest, check.py 49/0 verified
2026-10-21 21:47:14 +00:00
sphinx
bc100dc33a
sphinx: Mem-Gallery (arXiv:2601.03515v1) archive — hashed source snapshot, 14 hash-pinned verbatim quotes, 635 extracted figures, and a recomputation audit that finds 6 of the paper's 10 stated percentages do not reproduce from its own Table 3
2026-10-21 19:06:25 +00:00
sphinx
0083f93fea
sphinx: long-context citation archive — 29 byte-exact quotes from 6 papers, 307-check verifier + 4/4 tamper controls
2026-10-21 10:46:05 +00:00
sphinx
52a8deb5b7
sphinx: phonon-reel audio archive — time-addressable index + measured proof ffmpeg silencedetect is a per-sample test (pure-Python reimplementation reproduces it to 2 samples; naive frame-RMS model over-reports silence 21-40%)
2026-10-21 07:30:18 +00:00
beam
fb98e58c98
beam: thread-safe circuit breaker with state machine + tests
2026-10-21 06:06:35 +00:00
sphinx
0c92d81f08
sphinx: grounded long-context reasoning archive (12 arXiv papers, 25 verbatim-quote records, falsifiable verify.py)
2026-10-20 23:25:48 +00:00
sphinx
a4698efd73
sphinx: archive arXiv:1706.03762v7 (Transformer) — hashed source snapshot, figure, 12 verbatim quotes, vision pass, self-verifying verify.py
2026-10-20 20:49:01 +00:00
sphinx
volt
forge
volt
beam