ds006239 (roi-language): alignment results (control gate: fail)

roi-language/README.md

@@ -4,7 +4,7 @@ Wang et al. 2025 — word-level phonological and semantic reading tasks in child
 
 - Paper: https://www.sciencedirect.com/science/article/pii/S2352340925009692
 - Data: https://openneuro.org/datasets/ds006239/versions/1.0.5
-- Generated: 2026-09-10
+- Generated: 2026-09-11
 - Pipeline: https://github.com/suchirsalhan/cdl-representations-brains-babylms
 - Masking: **roi-language** -- see DATASETS.md section 10 for the three-level standard (phonology/language/all) this is part of, and how it differs from the whole-brain reference.
 
@@ -62,7 +62,7 @@ rather than language) and an inter-subject noise ceiling.
 | SemLocal | ses-11+   |       48 |        0.310001 |        0.359958 |          49 |
 | SemLocal | ses-11    |       48 |        0.26477  |        0.331669 |          34 |
 
-Model grid: **15 families**, 2096 alignment rows across 8 cells.
+Model grid: **15 families**, 2088 alignment rows across 8 cells.
 
 | | |
 |---|---|
@@ -79,7 +79,7 @@ Model grid: **15 families**, 2096 alignment rows across 8 cells.
 | pico-decoder-tiny     |              21 |     0.0053 |   0.0058 |        0.0538 |                    0.119  |               0      |
 | pico-decoder-small    |              21 |     0.0035 |   0.0125 |        0.0522 |                    0.1971 |               0      |
 | babylm-gpt2           |               9 |     0.002  |   0.0104 |        0.0389 |                    0.0997 |               0      |
-| pico-decoder-large    |              21 |     0.0015 |   0.009  |        0.0474 |                    0.162  |               0      |
+| pico-decoder-large    |              20 |     0.001  |   0.0089 |        0.0474 |                    0.162  |               0      |
 | pico-decoder-medium   |              21 |    -0.0007 |   0.0086 |        0.0476 |                    0.1798 |               0      |
 | beetle-humanscale-eng |              18 |    -0.0014 |   0.0122 |        0.0423 |                    0.1365 |               0      |
 | pythia-70m-full       |              21 |    -0.0016 |   0.0103 |        0.0382 |                    0.1389 |               0      |

roi-language/alignment_by_cell.csv

@@ -47,14 +47,14 @@ beetle-humanscale-eng,Sem,ses-11,18,-0.013188074771802928,0.02143256618349877,-0
 beetle-humanscale-eng,Sem,ses-11+,18,-0.004858015157744306,0.020698827637789012,-0.0214792062295682,0.0362764238987694,0.3901386367174326,,0.09298341790496255,
 beetle-humanscale-eng,SemLocal,ses-11,18,-0.010799362716976617,0.020424953308517933,-0.0280399981933714,0.0339337140240508,0.2647704592119395,,0.12816276455104098,
 beetle-humanscale-eng,SemLocal,ses-11+,18,-0.018390538005947678,0.026343225432008188,-0.0423209417357121,0.0267881170063189,0.3100011639992813,,0.08641295619903222,
-pico-decoder-large,Orth,ses-11,21,-0.006382505062119976,0.017754411524268515,-0.03486546192463,0.0226487988974522,0.5256372477199311,,0.04308826856486373,
-pico-decoder-large,Orth,ses-11+,21,-0.0049190168587500095,0.01810607771922322,-0.0289697587881934,0.0422497406913287,0.5653630359262645,,0.07473028480206297,
-pico-decoder-large,Phon,ses-11,21,-0.006382505062119976,0.017754411524268515,-0.03486546192463,0.0226487988974522,0.5256372477199311,,0.04308826856486373,
-pico-decoder-large,Phon,ses-11+,21,-0.0049190168587500095,0.01810607771922322,-0.0289697587881934,0.0422497406913287,0.5653630359262645,,0.07473028480206297,
-pico-decoder-large,Sem,ses-11,21,0.0015524709914889903,0.01662610386190612,-0.0210139880252401,0.0474267552049968,0.3510289007265154,,0.13510783615491168,
-pico-decoder-large,Sem,ses-11+,21,0.0030012132150487,0.012764357437463385,-0.0136153364328022,0.0312489408677112,0.3901386367174326,,0.08009701661602926,
-pico-decoder-large,SemLocal,ses-11,21,0.014716111732974062,0.017502583054116016,-0.0220010781913468,0.0428896601637125,0.2647704592119395,,0.16198808693148364,
-pico-decoder-large,SemLocal,ses-11+,21,0.015087526283690086,0.011233550706072806,-0.0030563274964662,0.0357534273811339,0.3100011639992813,,0.11533320365602495,
+pico-decoder-large,Orth,ses-11,20,-0.006683489395994055,0.018160590809398493,-0.03486546192463,0.0226487988974522,0.5256372477199311,,0.04308826856486373,
+pico-decoder-large,Orth,ses-11+,20,-0.005704472215856305,0.018205676034988447,-0.0289697587881934,0.0422497406913287,0.5653630359262645,,0.07473028480206297,
+pico-decoder-large,Phon,ses-11,20,-0.006683489395994055,0.018160590809398493,-0.03486546192463,0.0226487988974522,0.5256372477199311,,0.04308826856486373,
+pico-decoder-large,Phon,ses-11+,20,-0.005704472215856305,0.018205676034988447,-0.0289697587881934,0.0422497406913287,0.5653630359262645,,0.07473028480206297,
+pico-decoder-large,Sem,ses-11,20,0.0013626991356891246,0.017034672045861773,-0.0210139880252401,0.0474267552049968,0.3510289007265154,,0.13510783615491168,
+pico-decoder-large,Sem,ses-11+,20,0.003098552815725555,0.013087955324056273,-0.0136153364328022,0.0312489408677112,0.3901386367174326,,0.08009701661602926,
+pico-decoder-large,SemLocal,ses-11,20,0.013585387385281606,0.017152293847102228,-0.0220010781913468,0.0428896601637125,0.2647704592119395,,0.16198808693148364,
+pico-decoder-large,SemLocal,ses-11+,20,0.014744101957391004,0.011411716778457193,-0.0030563274964662,0.0357534273811339,0.3100011639992813,,0.11533320365602495,
 pico-decoder-medium,Orth,ses-11,21,-0.00819581117379288,0.01833653377870844,-0.0376405653034201,0.0289485256597728,0.5256372477199311,,0.055073200739375855,
 pico-decoder-medium,Orth,ses-11+,21,-0.0035681912421515524,0.017832293176091175,-0.0310112614084039,0.0388038529498783,0.5653630359262645,,0.0686352847357696,
 pico-decoder-medium,Phon,ses-11,21,-0.00819581117379288,0.01833653377870844,-0.0376405653034201,0.0289485256597728,0.5256372477199311,,0.055073200739375855,

roi-language/alignment_by_checkpoint.csv

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:68974ad5a51b437c257fc1f6c64d55e750f6d3e6b62f18844a50dabee65ab038
-size 553380
+oid sha256:85d52a3c0fa8e65a1a062cff23278f3bc99dd80960b3e6c8820b8d00ef9d9e6a
+size 551132

roi-language/alignment_by_family.csv

@@ -2,7 +2,7 @@ family,params,n_checkpoints,n_rows,n_cells,rsa_mean,rsa_sd,rsa_sd_all_rows,rsa_a
 pico-decoder-tiny,11282784.0,21,168,8,0.005251294788689426,0.005765810379518547,0.018298350465113408,0.0538436152749142,0.0538436152749142,Orth/ses-11+,30000,0.010440014867050127,0.11902850253083871,,,,,False,5.140438390786636e-08,True,
 pico-decoder-small,64595328.0,21,168,8,0.0035338403553263112,0.012539751513951583,0.018864863004069057,0.05218683234362,0.05218683234362,SemLocal/ses-11,5000,0.01685415190524445,0.1971021710614864,,,,,False,7.842838865682486e-06,True,
 babylm-gpt2,134601216.0,9,72,8,0.0019970468267228292,0.010354606992853975,0.014464454852087276,0.038879960106291,0.038879960106291,Sem/ses-11+,598,0.007371562632699089,0.09965677953206864,,,,,False,1.7500429581915914e-06,True,
-pico-decoder-large,569808384.0,21,168,8,0.0014692847976827325,0.0090096032396843,0.01816296412601773,0.0474267552049968,0.0474267552049968,Sem/ses-11,5000,0.009334889442851222,0.16198808693148364,,,,,False,6.318255982825595e-07,True,
+pico-decoder-large,569808384.0,20,160,8,0.0010018522587983212,0.008942671983029034,0.018284232607333934,0.0474267552049968,0.0474267552049968,Sem/ses-11,5000,0.00813572441248982,0.16198808693148364,,,,,False,5.6248990639436e-07,True,
 pico-decoder-medium,181095168.0,21,168,8,-0.0007405860083300553,0.008610867813134157,0.01835642842794999,0.0476085825288689,0.0476085825288689,SemLocal/ses-11,3000,0.002962656088356319,0.17981077900673162,,,,,False,4.1896017205180245e-07,True,
 beetle-humanscale-eng,193811712.0,18,144,8,-0.0014000188992872586,0.01216257235214028,0.021672335958979598,0.0423209417357121,-0.0423209417357121,SemLocal/ses-11+,64,-0.01064211103793033,0.1365186542841827,,,,,False,4.748218928099878e-06,True,
 pythia-70m-full,95592496.0,21,168,8,-0.001618431060929435,0.010302304259638118,0.017642731632360194,0.0381892478191561,-0.0381892478191561,Orth/ses-11+,43000,0.0023710636411192754,0.13891946634483374,,,,,False,1.6036890683094196e-06,True,

roi-language/seed_null_comparison.csv

@@ -2,7 +2,7 @@ family,rsa_mean,rsa_sd,mean_in_seed_sd,rsa_abs_max,null_max_p50,null_max_p95,bes
 pico-decoder-tiny,0.005251294788689426,0.005765810379518547,,0.0538436152749142,,,,False,0.11902850253083871,5.140438390786636e-08,True
 pico-decoder-small,0.0035338403553263112,0.012539751513951583,,0.05218683234362,,,,False,0.1971021710614864,7.842838865682486e-06,True
 babylm-gpt2,0.0019970468267228292,0.010354606992853975,,0.038879960106291,,,,False,0.09965677953206864,1.7500429581915914e-06,True
-pico-decoder-large,0.0014692847976827325,0.0090096032396843,,0.0474267552049968,,,,False,0.16198808693148364,6.318255982825595e-07,True
+pico-decoder-large,0.0010018522587983212,0.008942671983029034,,0.0474267552049968,,,,False,0.16198808693148364,5.6248990639436e-07,True
 pico-decoder-medium,-0.0007405860083300553,0.008610867813134157,,0.0476085825288689,,,,False,0.17981077900673162,4.1896017205180245e-07,True
 beetle-humanscale-eng,-0.0014000188992872586,0.01216257235214028,,0.0423209417357121,,,,False,0.1365186542841827,4.748218928099878e-06,True
 pythia-70m-full,-0.001618431060929435,0.010302304259638118,,0.0381892478191561,,,,False,0.13891946634483374,1.6036890683094196e-06,True

roi-language/summary.json

@@ -2,7 +2,7 @@
   "dataset": "ds003604",
   "rdms": "within-run normalised",
   "n_families": 15,
-  "n_rows": 2096,
+  "n_rows": 2088,
   "n_cells": 8,
   "ceiling_lower_mean": 0.43724246599344496,
   "best_rsa_abs": 0.0690762309770627,
@@ -14,13 +14,13 @@
   "sesoi": 0.05,
   "scale_trend_rho": -0.34779317689787076,
   "scale_trend_p": 0.027874566812770417,
-  "var_pct_cell": 7.468620429390696,
-  "var_pct_family": 16.527850065877438,
-  "var_pct_residual": 75.95581958106771,
+  "var_pct_cell": 7.4672226073819346,
+  "var_pct_family": 16.440164175981923,
+  "var_pct_residual": 76.04472049632966,
   "untrained_mean": -0.00047131131310623113,
   "untrained_n": 88,
-  "trained_mean": -0.005947320002564258,
-  "trained_n": 2008,
-  "untrained_vs_trained_t": 2.355518686072725,
-  "untrained_vs_trained_p": 0.020543072021083426
+  "trained_mean": -0.006014381024875999,
+  "trained_n": 2000,
+  "untrained_vs_trained_t": 2.384122787302241,
+  "untrained_vs_trained_p": 0.019097154321923073
 }

roi-language/training_trend.csv

@@ -12,5 +12,5 @@ pythia-160m-full,8,-0.04774277614581618,0.6743690818335736,-0.20024192493278836,
 pythia-1.4b-full,8,-0.11286132456238866,0.3863319424291648,-0.8262843339327903,0.4358974364859556
 pico-decoder-small,8,-0.12987012987012986,0.2590966050866248,-1.4177267891880536,0.1992123167581929
 pico-decoder-medium,8,-0.2383116883116883,0.31561228184229845,-2.135681284106805,0.07007980431773414
-pico-decoder-large,8,-0.39464285714285713,0.2879968173978897,-3.8758017252254957,0.006085892144156609
 pythia-410m-full,8,-0.3980188582768554,0.556354355258752,-2.02347177526307,0.08270560591041042
+pico-decoder-large,8,-0.44473684210526304,0.283535989357202,-4.436494112921112,0.0030204611184561064