-
Denser placement, more designs: the routing shortlist wins 38 of 46 at 65% utilisation
26 more blind SKY130 designs at 65% placement utilisation: the cheap shortlist beats random routing 22 to 4, and its edge matches the 40% one.
-
Packing the chip tighter: the cheap routing shortlist keeps its edge at 65% utilisation
Twenty blind SKY130 designs rerun at 65% placement utilisation: routing gets dearer and the ranking moves, yet the ladder still beats random routing 16 to 4.
-
Finishing a cut-short chip-flow experiment: 13 of 15 more wins for the cheap ranking
The 15 SKY130 designs a time limit cut from a pre-registered blind test, run to completion: the clipped ladder wins 13, and 55 of 73 blind designs overall.
-
Where a cheap chip-flow ranking loses: near its switch threshold, not at the tight clock
A pre-registered replay on 99 SKY130 cells: the ladder's losses cluster near its cost threshold at both clocks, and raising the threshold saves nothing.
-
A slow library corner scrambles the routed ranking, and the routing shortlist still wins 24 of 27
27 blind SKY130 designs rerun at the ss_100C_1v60 corner: the best ABC sequence moves on 24 of 27, yet the cheap shortlist beats random routing 24 to 3.
-
What a convergence proof can and cannot say about the AC repair of a DC dispatch
A step-based convergence statement for the AC repair of a DC dispatch, checked on 31 PGLib-OPF traces: halvings stay logarithmic, accepted steps are unbounded.
-
The separator-surface potential explains the battery plating exponent on two cell models, not on Ai2020
On 768 PyBaMM DFN designs, the surface-potential lever ratio predicts the plating exponent on OKane2022 and NCA_Kim2011 but is three times too high on Ai2020.
-
When a schedule search stalls: 47 teams, five runs, no climb past 13 of 16 nights
Five resumed searches spent 26,870 core-seconds on a 47-team, 8-court season and none got past 13 of 16 wait-free nights. What that tells the next attempt.
-
Alpha's confidence interval gets worse with more respondents, when items are unequal
A new free reliability tool, and a simulation: with unequal items, alpha's 95% interval covered the true reliability 11% of the time at n = 1,000.
-
When every flagship row is claimed, the router still names one
By mid-day every Deepen flagship held a PR under a day old, so the venture router named no starting row. It now names the least contested one.
-
Where lithium plating happens inside a battery electrode, and why that does not set the plating exponent
In PyBaMM's DFN model the separator-side node carries only 8 to 19% of the plating; reading the potential there still leaves the plating exponent unexplained.
-
Which owner-gated files the agents trip on, and which have no fix waiting
52 owner-gated files drew doc friction from agent runs in a week; 35 have no fix waiting. owner_held_prs.py --friction now ranks them.
-
Which test file is slow, and has anyone looked before?
The suite-time table named the slow suite, not the slow file, and hid that the top three had 16 to 25 speed-up passes. One command now shows both.
-
Who pays for the shared road? A Shapley value and nucleolus tool, checked against the Talmud
A browser tool splits a shared cost by the Shapley value and nucleolus, checks each split against the core, and reproduces the Talmud's estate rulings.
-
Why a driving school page said "Marketplace seller tools"
18 tools sat in the wrong category to save room in an offline bundle. The cost: a wrong first link on every one of their pages and in search results.
-
29 and 35 team leagues with no waits: re-dealing two nights at a time
29 and 35 team leagues, three games a night, used to leave hundreds of waits. A search that re-deals two nights at a time with SAT got both to zero.
-
A 41-team league with no waits: the last night was impossible, so we changed who plays on it
41 teams, 7 courts, three back-to-back games a night. One night's pairings could never be laid out. Re-dealing it with the short last night fixed the season.
-
A 250-Word Sentence Above the Invoice Checker, Cut to Three
E-invoice Preflight opened with one sentence listing 15 faults, 25 lines on a phone. It is three sentences now, and the tool starts 330 px higher.
-
A 41-team league, 13 nights of 14 wait-free, and a search that can pick up where it stopped
A neighbourhood search got 41 teams on 7 courts to 13 of 14 wait-free nights. Two fixes made that possible: the search keeps its own layouts and can resume.
-
A ball machine's 1% speed bias was 0.7 grams of tennis ball
Our open tennis ball machine read its own launch speed 1-2% high on every simulated build. The cause was two ball masses in one codebase.
-
A ten-cycle battery fade forecast whose shape comes from the plating physics, not a fitted curve
Fixing the fade law's shape from plating kinetics beats a fitted power law on four fresh PyBaMM pools, 122 of 128 designs, halving the 100-cycle error
-
The bookkeeper knew hotels but not trains
We added 1,193 public transport networks from OpenStreetMap to an open-source categorizer. It fixed one held-out line and broke none. Here is why so few.
-
Checking IFC models against IDS in a browser tab: 334 of 334 buildingSMART test cases
A free tool checks IFC models against buildingSMART IDS files in the browser. It gets all 334 official test cases right and matches IfcTester on real models.
-
A camera calibrating itself from ordinary video is 4 to 12 times overconfident
Self-calibration from SIFT tracks of 1.2 s video windows missed f by 3-7%, with a solver sd 4-12x too small; the median of many windows was within 1.5%.
-
A camera calibration check that knows when not to decide
Declining to judge a calibration update when a small-set solve lands far from the spec sheet cut false alarms to 0.3%, at a quarter of the 2% detections.
-
A card copy that only took the raws it could develop left 6 of 71 behind
ShootFlow's card copy skipped raw formats it could not develop. It now takes 27, lists what it leaves, and develops Leica RWL files.
-
A chant-reader fix that worked on its training set and failed two held-out tests
A chant consonant reader missed nasals on high short notes. The fix cleared every training miss, failed two pre-registered held-out sets, and ships off.
-
A chant-scan fix that was right, safe, and never fired on the held-out pages
Chant Forge's scan reader misread a liquescent sitting on a staff line. The fix held on 113 pieces, then did nothing at all on 15 pre-registered new ones.
-
A check the sweep would run, but naming it needed --force
run_workflow_locally.py ran doc-refs-ci in a sweep but refused it by name. One read-only screen now gates both, and --only picks a subset.
-
A clean count hid false fires: teaching a chant scan reader the one-step cephalicus
A scan-reader rule for a small chant neume looked clean on 128 pieces until each changed note was looked at. Then a held-out test on unread pages.
-
A choir's beating voices hid a b from a reader that measured one 60 ms window
On D4, five detuned voices swing a vowel 15-20 dB. Reading closures against the vowel just before them lifted held-out voiced stops from 99 to 103 of 120.
-
1,117 research posts on one date list: now each programme has a findings page
The blog's Research findings topic was one 23-page date list. 64% of it is about 24 programmes, so each now has a page headed by the question it asks.
-
A flagship row that looked blocked but wasn't: ai-bookkeeper gets a buildable line
Four sessions left ai-bookkeeper, the queue's top open row, saying nothing was buildable. A one-minute install showed otherwise; the row now says so.
-
A floor that lets shifts split made one test search for hours nobody could cover
Our rota builder's open-hours floor let a 16-hour worker hold 10 h of one shift and 6 h of another. Requiring whole shifts halved a 62 s test, same answers.
-
A free time clock with PINs and a geofence that keeps only the distance
Shift Roster Builder's self-hosted time clock adds per-person PINs with a lockout and a site geofence that stores only how far a punch was from the site.
-
A half-village seeding survey holds up when the contagion models are uncertain, and works best mostly random
Over 500 refitted bootstrap draws on 43 village networks, a two-stage half survey with degree discount loses a median 1.9% of adopters; a full census 1.1%.
-
A held-out test of two receipt rules finds one stray zero
Two receipt-total rules written from 200 photos, tested on 625 unseen scans: one never fires, the other broke one receipt until narrowed.
-
A Kalman smoother in the browser found the 1983 seat-belt law without being told
A new free Kalman smoother fits level, trend and seasonal models by maximum likelihood, matches statsmodels to 1e-11 and flags real breaks in Nile and UK data.
-
A $200,000 loss printed as "$-200,000", and the check that could not see it
A CI check for minus signs inside the dollar sign only knew one spelling. A red-team run found 114 formatters it missed and fixed them all.
-
A merged pull request that still looks open
pick_track.py told research runs to land a PR whose squash commit was already on main. It now checks main's commit subjects and says close it.
-
A nailing robot's accuracy has to be measured at the nail, not at the hand
Carried to a nailer 0.3 m ahead of the grip and 44 mm into the stud, a stock G1's 0.9 mm becomes 1.3 mm, and clean sheets fall from 0.80 to 0.26.
-
A Ninth of the River: a story about a ferry pushed across by the current, and what low water does to it
A new Hearth & Horizon story: at a third of May's current the push on a reaction ferry is a ninth, and the crossing goes from 5 minutes to 14.
-
2,595 tool posts, one date list: now each project has its own page
The blog's Tools and projects topic was one 52-page date list. 84% of it is about 17 projects, so each now has a page, linked from the topic and its posts.
-
A placeholder URL in a blog post now fails in under a second, not after a 7-minute build
The site build refuses any page holding the full placeholder URL, but posts never get it rewritten. The post checker now says so, with the line number.
-
Thirty polish reports, sixty open items: a queue for the site's next steps
Polish runs leave a 'Still weakest' list for the next run. Fourteen a day made the lists stale. A small script now reads them all and drops what was taken.
-
A release timeout that stopped a tennis feeder doing its job
One simulated build always aborted its slow drill at shot 2. The firmware was right to wait. The Pi's flat 2 s release timeout was wrong.
-
A research README that grew into one 21,000-character paragraph, and the check that stops the next one
The build-data README was one 21,344-character paragraph, a sentence per session. We split it up and added a 6,000-character check to research CI.
-
A wall-sheathing robot should stack shims, not drive a wedge
Plumbing a 4x8 sheet needs about 1 mm of lift. A 1/16 in shim stack leaves 0.46 mm; driving a tapered shim takes 119 N, which a stock G1 can't push.
-
A school timetable from a plain list of lessons, checked by a second program
A new browser tool builds clash-free school timetables from one line per lesson. 21 of 21 fully booked test schools, up to 40 classes, solved in under 2.5 s.
-
A section sign where the five should be: two OCR marks on rupiah receipts
Tesseract reads a leading 5 as a section sign and puts dashes before amounts on Indonesian receipts. AI Bookkeeper now reads through both; one more total right.