queue: land the three gated filings; let the dispatch door run Claude

.github/workflows/queue_filing.yml

@@ -202,6 +202,14 @@ jobs:
           claude_code_oauth_token: ${{ secrets.CLAUDE_CODE_OAUTH_TOKEN }}
           # Default-hidden output masks real failures — surface everything.
           show_full_output: true
+          # The dispatch door (queue_actions.yml's hand-over, the nightly
+          # sweep) runs as github-actions, and the action refuses a run
+          # "initiated by non-human actor" unless that bot is allowed — the
+          # first sweep's three filings all died on exactly that line
+          # (2026-09-10). The `labeled` door is still the human's tap; the
+          # author-association gate above is what keeps strangers out on
+          # both doors.
+          allowed_bots: "github-actions"
           # Workspace edits + the repo's own gates + arXiv metadata lookups.
           # No git/gh: the deterministic steps below own branch/commit/PR.
           claude_args: '--allowedTools "Read,Glob,Grep,Edit,Write,Bash(make:*),Bash(python:*),Bash(python3:*),WebFetch(domain:arxiv.org),WebFetch(domain:export.arxiv.org)"'
@@ -263,16 +271,19 @@ jobs:
                  place, exactly as the `reading-queue.md` case above. If it
                  is already DONE-marked anywhere in reading-queue.md, the
                  paper has been filed before — stop and report that, without
-                 editing anything.
+                 editing anything. If it is in NONE of the three files, the
+                 suggestion simply lapsed before the tap was acted on: the
+                 request itself carries the title and ref, so file it
+                 exactly as the suggestion-tier case above — append the DONE
+                 line to the section — with nothing left to delete.
             5. Run `make validate` and `python -m pytest tests/ -q`; fix your
                own filing mistakes until both are green.
 
             Touch ONLY bibliography/, wiki/, reading-queue.md,
             arxiv-inbox.md and arxiv-interests.md. Do not
             commit — the workflow handles git. If the paper cannot be
-            identified confidently, or its line cannot be found in any of
-            the three files, write the reason to /tmp/filing_error.txt and
-            stop without editing anything.
+            identified confidently, write the reason to
+            /tmp/filing_error.txt and stop without editing anything.
 
       - name: Abort cleanly if Claude could not identify the paper
         if: steps.resolve.outputs.skip != 'yes' && steps.dup.outputs.skip != 'yes'

arxiv-inbox.md

@@ -29,7 +29,6 @@ Format:
 ---
 last digest: 2026-09-10
 2026-09-07 — Probing Dark Matter with Strongly Lensed Binary Black Hole Mergers: Prospects in the Near Future — 2609.05020
-2026-09-07 — ALPACA I: Controlling source and PSF systematics in JWST time-delay cosmography with differentiable lens modeling — 2609.04312
 2026-09-09 — Harnessing stellar kinematics to constrain dark energy with the double-source-plane gravitational lens SDSS J0946+1006 — 2609.08573
 2026-09-09 — $\texttt{LensFactory.jl}$: A general-purpose strong lens modeling package — 2609.08649
 2026-09-09 — Observation-driven simulations of strong lensing galaxy clusters — 2609.07840

bibliography/pyautomemory.bib

@@ -16127,6 +16127,16 @@ @unpublished{Vincken2026
   note = {In preparation; arXiv ID, DOI and full author list TODO once public}
 }
 
+@misc{Adam2026,
+  archivePrefix = {arXiv},
+  author = {Adam, Alexandre},
+  eprint = {2608.18224},
+  primaryClass = {astro-ph.CO},
+  title = {{Mapping the Information Geometry of an Unresolved Dark Matter Population using a Differentiable Strong Lensing Simulator}},
+  url = {https://arxiv.org/abs/2608.18224},
+  year = {2026}
+}
+
 @article{Morgan2026,
   archivePrefix = {arXiv},
   arxivId = {2608.17041},
@@ -16139,3 +16149,21 @@ @article{Morgan2026
   url = {https://arxiv.org/abs/2608.17041},
   year = {2026}
 }
+
+@article{Goobar2026,
+  archivePrefix = {arXiv},
+  arxivId = {2608.28413},
+  author = {Goobar, Ariel and Johansson, Joel and M{\"{o}}rtsell, Edvard and Lemon, Cameron and Schulze, Steve and Dhawan, Suhail and Townsend, Alice and Li, Maggie L. and Qin, Yu-Jing and Yan, Lin and Fremling, Christoffer and Kasliwal, Mansi M. and Nugent, Peter and Smith, Graham P. and Sollerman, Jesper and Andreoni, Igor and Arendse, Nikki and Collett, Thomas E. and Nordin, Jakob and Hjortlund, Jacob Osman and Perley, Daniel A. and Rigault, Mickael and Thorp, Stephen and Wise, Jacob L.},
+  doi = {10.48550/arXiv.2608.28413},
+  eprint = {2608.28413},
+  note = {Submitted to the ApJ Focus Issue on SN 2025wny (Paper I)},
+  title = {{Follow-up of SN 2025wny I: Space-based Observations of the First Multiply-imaged Superluminous Supernova}},
+  url = {https://arxiv.org/abs/2608.28413},
+@misc{Krizic2026,
+  author = {Krizic, Hrvoje and Millon, Martin and Galan, Aymeric and Erickson, Sydney and Marshall, Phil},
+  eprint = {2609.04312},
+  primaryClass = {astro-ph.IM},
+  title = {{ALPACA I: Controlling source and PSF systematics in JWST time-delay cosmography with differentiable lens modeling}},
+  url = {https://arxiv.org/abs/2609.04312},
+  year = {2026}
+}

reading-queue.md

@@ -86,12 +86,14 @@ High-Dimensional Bayesian Model Comparison in Cosmology with GPU-accelerated Nes
 THE STELLAR AND DARK MATTER DISTRIBUTIONS IN ELLIPTICAL GALAXIES MEASURED BY  STACKED WEAK GRAVITATIONAL LENSING
 Model independent lensing sub-structure detection with multiply-imaged star clusters constellations — 2608.21253
 On what scales does strong lensing robustly constrain the mass profile of low-mass perturbers? — 2608.20454
-Follow-up of SN 2025wny I: Space-based Observations of the First Multiply-imaged Superluminous Supernova — 2608.28413
 Measuring Extragalactic Microlens Masses and Motions in Strongly Lensed Quasar Systems with Intensity Interferometry — 2608.27577
 TDCOSMO XXIX: JWST/NIRSpec IFU Spatially Resolved Kinematics of Three Time-delay Lenses — 2609.03089
 Host Dependence and Line-of-Sight Effects on Galaxy-Galaxy Strong Lensing in Clusters — 2608.30079
 TDCOSMO XXVIII. The Hubble constant from the quadruply lensed quasar J1537$-$3010 with precise time delays — 2609.00179
 HOLISMOKES -- XVIII. Cosmology with strongly lensed type II supernovae: Effects of instrumental setups on $H_0$ — 2608.26331
+DONE 2026-09-01 — Mapping the Information Geometry of an Unresolved Dark Matter Population using a Differentiable Strong Lensing Simulator — 2608.18224
+DONE 2026-09-03 — Follow-up of SN 2025wny I: Space-based Observations of the First Multiply-imaged Superluminous Supernova — 2608.28413
+DONE 2026-09-07 — ALPACA I: Controlling source and PSF systematics in JWST time-delay cosmography with differentiable lens modeling — 2609.04312
 
 
 ## SMBHs

wiki/lensing/sources/dark-matter-substructure.md

@@ -523,6 +523,22 @@ Per-paper stubs of substructure-detection literature. All marked stub.
 - Evidence until the paper identity and canonical key are verified.
 
 
+## Adam 2026 — Fisher information geometry of subhalo populations
+
+**Canonical BibTeX key:** `Adam2026`
+**Reference:** Mapping the Information Geometry of an Unresolved Dark Matter Population using a Differentiable Strong Lensing Simulator; arXiv:2608.18224
+**Concepts:** [[dark-matter-substructure]], [[multipoles]], [[line-of-sight-effects]], [[mass-sheet-degeneracy]]
+
+**Supports:**
+- Represents an unresolved subhalo population with a spectral basis carrying both radial and angular structure, then uses a differentiable strong-lensing simulator and Fisher matrix analysis to quantify how much macro-model and source-model flexibility erode substructure sensitivity.
+- Proposes a Fisher Graph Laplacian prior as a diagnostic for managing the competing macro/source/substructure model components — a linear-algebra approach worth reusing for population-level subhalo fitting.
+
+**Use when:**
+- Citing a Fisher-information / linear-algebra treatment of macro- and source-model degeneracies for an unresolved subhalo population, or its radial+angular spectral-basis parameterization.
+
+**Do not use for:**
+- Claims about how line-of-sight structure, external multipoles, or the mass-sheet transform specifically interact with the radial subhalo signal — the paper does not settle this; it remains an open question to check against the text before citing.
+
 ## See also
 
 - [[dark-matter-substructure]]

wiki/lensing/sources/lensed-supernovae.md

@@ -41,6 +41,10 @@ status: drafted
 **Do not use for:**
 - An observed lensed-supernova sample or forecasts for Roman or JWST.
 
+## Goobar et al. 2026 — SN 2025wny I, space-based observations
+
+**Canonical BibTeX key:** `Goobar2026`
+
 ## See also
 
 - [[lensed-supernovae]]

wiki/lensing/sources/time-delay-cosmography.md

@@ -137,6 +137,25 @@ status: stub
 - Evidence until the paper identity and canonical key are verified.
 
 
+## Krizic 2026 — ALPACA I (JWST PSF/source systematics)
+
+**Canonical BibTeX key:** `Krizic2026`
+**Reference:** ALPACA I: Controlling source and PSF systematics in JWST time-delay cosmography with differentiable lens modeling; arXiv:2609.04312
+**Concepts:** [[time-delay-cosmography]], [[source-reconstruction]]
+
+**Supports:**
+- Demonstrates ALPACA, a differentiable forward-modeling pipeline that jointly reconstructs source-galaxy morphology, lens mass distribution, and PSF for JWST-observed lensed quasars, reducing H0 bias from ~14% to ~1% (on mock data) relative to fixed/mismodeled-PSF fitting.
+- Describes a concrete recipe for PSF modelling of JWST lensed quasars that is a candidate to build into PyAutoReduce's reduction pipeline.
+
+**Use when:**
+- Citing PSF systematics as a leading bias source in JWST time-delay H0 measurements, and joint source+PSF+mass modelling as the mitigation.
+- Referencing differentiable lens-modeling pipelines for JWST time-delay cosmography, or JWST PSF-modelling approaches relevant to PyAutoReduce.
+
+**Do not use for:**
+- Claims about results from real (non-mock) JWST time-delay lens systems — the headline bias-reduction figures are demonstrated on simulated data.
+- Non-JWST PSF systematics or non-time-delay lens-modelling contexts without checking applicability.
+
+
 ## See also
 
 - [[time-delay-cosmography]]