Official POLAR Shared Task Results

Participants Results

This page summarizes the official POLAR @ SemEval-2026 results for Subtasks 1-3. The complete F1 Macro rankings are available on the dedicated leaderboards page.

View Full Leaderboards Official Leaderboards Repository
3Official subtasks covered on this results page.
22Languages represented across the official ranking files.
62Language-specific leaderboard tables available on the leaderboards page.
1529Ranked official team entries across all tables.
69Total official teams represented across the leaderboards.

Top-3 Placement Summary

These tables count how often each team appears in the top three positions across the language-specific leaderboards for each subtask.

Subtask 1: Polarization Detection

TeamTotal1st2nd3rd
UTokyo Tsuruoka Lab12840
NYCU-NLP12345
PSK8323
Lingo Research Group5122
SMASH5122
taien3111
yunkuang03293102
OZemi2101
StanceLab2020
MKJ2011
Tralaleros2002
mdok-style1100
PhatThachDau1100
Sagarmatha1100
CUET-8231010
danielkhir1010
JAT1010
Projet Fil Rouge 8211010
PolaFusion1001
Semantic Vectors1001
UMUSP1001

Subtask 2: Polarization Type Classification

TeamTotal1st2nd3rd
NYCU-NLP14554
SMASH14446
UTokyo Tsuruoka Lab12651
Lingo Research Group6213
CoPol5320
Sagarmatha3012
NASIM LAB2101
PolaFusion2101
AIvengers1010
ILA Polar1010
maggam1010
Stochastic Gradient Descenders1010
mdok-style1001
MSqrd1001
YNU-HPCC1001
yunkuang03291001

Subtask 3: Manifestation Identification

TeamTotal1st2nd3rd
SMASH16943
NYCU-NLP11731
PolaFusion7025
Sagarmatha4202
happynewyear4040
AIvengers4004
OZemi2020
YEZE2002
Aaronbundi1010
maggam1010
suuii1010
Lingo Research Group1001

Top Three Systems by Language

Each panel lists the top three official systems for every language, with the POLAR baseline included for quick comparison.

Top 3 + baseline
Subtask 1: Polarization Detection 22 languages

amh

TeamScore
PSK0.8002
UTokyo Tsuruoka Lab0.7954
Lingo Research Group0.7928
baseline0.7151

arb

TeamScore
UTokyo Tsuruoka Lab0.8488
PSK0.8484
NYCU-NLP0.8427
baseline0.7957

ben

TeamScore
UTokyo Tsuruoka Lab0.8625
CUET-8230.8582
NYCU-NLP0.8538
baseline0.8528

deu

TeamScore
NYCU-NLP0.7608
UTokyo Tsuruoka Lab0.7531
yunkuang03290.7465
baseline0.6714

eng

TeamScore
UTokyo Tsuruoka Lab0.8252
danielkhir0.8189
PSK0.8177
baseline0.7802

fas

TeamScore
baseline0.8424
OZemi0.8348
taien0.8314
MKJ0.8308

hau

TeamScore
PhatThachDau0.8336
Projet Fil Rouge 8210.8324
OZemi0.8313
baseline0.7753

hin

TeamScore
PSK0.8236
Lingo Research Group0.8212
Tralaleros0.8178
baseline0.7379

ita

TeamScore
mdok-style0.7303
baseline0.6773
StanceLab0.6720
PolaFusion0.6714

khm

TeamScore
SMASH0.7744
StanceLab0.7610
Semantic Vectors0.7553
baseline0.6592

mya

TeamScore
taien0.8913
MKJ0.8874
NYCU-NLP0.8868
baseline0.8210

nep

TeamScore
NYCU-NLP0.9236
Lingo Research Group0.9180
SMASH0.9136
baseline0.8798

ori

TeamScore
UTokyo Tsuruoka Lab0.8255
JAT0.8157
UMUSP0.8137
baseline0.7765

pan

TeamScore
UTokyo Tsuruoka Lab0.8257
PSK0.8121
NYCU-NLP0.8107
baseline0.7898

pol

TeamScore
Lingo Research Group0.8431
NYCU-NLP0.8350
PSK0.8348
baseline0.7241

rus

TeamScore
UTokyo Tsuruoka Lab0.8303
NYCU-NLP0.8232
yunkuang03290.8138
baseline0.7457

spa

TeamScore
UTokyo Tsuruoka Lab0.8030
NYCU-NLP0.7996
SMASH0.7976
baseline0.7266

swa

TeamScore
PSK0.8113
SMASH0.8098
taien0.7985
baseline0.7571

tel

TeamScore
Sagarmatha0.9053
SMASH0.9006
Tralaleros0.8968
baseline0.644

tur

TeamScore
NYCU-NLP0.8329
UTokyo Tsuruoka Lab0.8303
PSK0.8092
baseline0.6957

urd

TeamScore
UTokyo Tsuruoka Lab0.8196
NYCU-NLP0.8169
Lingo Research Group0.8156
baseline0.7890

zho

TeamScore
yunkuang03290.9315
UTokyo Tsuruoka Lab0.9289
NYCU-NLP0.9273
baseline0.8691
Subtask 2: Polarization Type Classification 22 languages

amh

TeamScore
PolaFusion0.6697
CoPol0.6579
SMASH0.6495
baseline0.3716

arb

TeamScore
NYCU-NLP0.6698
UTokyo Tsuruoka Lab0.6678
SMASH0.6581
baseline0.4855

ben

TeamScore
Lingo Research Group0.4216
NYCU-NLP0.4007
SMASH0.3780
baseline0.2887

deu

TeamScore
UTokyo Tsuruoka Lab0.6200
NYCU-NLP0.6157
Lingo Research Group0.5994
baseline0.4078

eng

TeamScore
UTokyo Tsuruoka Lab0.5322
Stochastic Gradient Descenders0.5157
NYCU-NLP0.5135
baseline0.3333

fas

TeamScore
SMASH0.6438
ILA Polar0.6207
MSqrd0.6088
baseline0.4626

hau

TeamScore
NYCU-NLP0.4796
SMASH0.4535
Sagarmatha0.4269
baseline0.2038

hin

TeamScore
SMASH0.8073
NYCU-NLP0.8013
YNU-HPCC0.7932
baseline0.7911

ita

TeamScore
UTokyo Tsuruoka Lab0.5505
maggam0.5375
yunkuang03290.4836
baseline0.3759

khm

TeamScore
UTokyo Tsuruoka Lab0.7048
SMASH0.7018
PolaFusion0.6986
baseline0.6268

mya

TeamScore
NASIM LAB0.7474
SMASH0.7358
UTokyo Tsuruoka Lab0.7079
baseline0.4772

nep

TeamScore
NYCU-NLP0.8104
Lingo Research Group0.8047
mdok-style0.8026
baseline0.7219

ori

TeamScore
UTokyo Tsuruoka Lab0.6027
AIvengers0.5938
NYCU-NLP0.5779
baseline0.5600

pan

TeamScore
SMASH0.5526
CoPol0.5258
Sagarmatha0.5243
baseline0.3650

pol

TeamScore
UTokyo Tsuruoka Lab0.6497
NYCU-NLP0.6400
Lingo Research Group0.6253
baseline0.4491

rus

TeamScore
CoPol0.6455
NYCU-NLP0.6295
SMASH0.6186
baseline0.5904

spa

TeamScore
NYCU-NLP0.6806
UTokyo Tsuruoka Lab0.6735
SMASH0.6726
baseline0.5935

swa

TeamScore
SMASH0.5694
UTokyo Tsuruoka Lab0.5397
NASIM LAB0.5360
baseline0.4417

tel

TeamScore
CoPol0.5734
Sagarmatha0.4647
SMASH0.4578
baseline0.3145

tur

TeamScore
CoPol0.7767
UTokyo Tsuruoka Lab0.6524
NYCU-NLP0.6462
baseline0.4708

urd

TeamScore
Lingo Research Group0.7978
SMASH0.7897
NYCU-NLP0.7889
baseline0.7127

zho

TeamScore
NYCU-NLP0.8436
UTokyo Tsuruoka Lab0.8350
Lingo Research Group0.8250
baseline0.6697
Subtask 3: Manifestation Identification 18 languages

amh

TeamScore
SMASH0.5789
NYCU-NLP0.5587
AIvengers0.5535
baseline0.4433

arb

TeamScore
NYCU-NLP0.6456
SMASH0.6413
YEZE0.6097
baseline0.3902

ben

TeamScore
SMASH0.2805
happynewyear0.2554
PolaFusion0.2493
baseline0.0868

deu

TeamScore
NYCU-NLP0.5176
maggam0.5153
SMASH0.5126
baseline0.3485

eng

TeamScore
Sagarmatha0.5105
happynewyear0.5071
SMASH0.5070
baseline0.4100

fas

TeamScore
SMASH0.4932
OZemi0.4764
Sagarmatha0.4611
baseline0.2004

hau

TeamScore
baseline0.7456
Sagarmatha0.2072
OZemi0.2058
PolaFusion0.2041

hin

TeamScore
SMASH0.7709
NYCU-NLP0.7704
PolaFusion0.7587
baseline0.2348

khm

TeamScore
baseline0.6095
SMASH0.4372
PolaFusion0.3998
AIvengers0.3774

nep

TeamScore
NYCU-NLP0.7127
SMASH0.7118
Lingo Research Group0.6685
baseline0.1314

ori

TeamScore
baseline0.3841
SMASH0.3296
happynewyear0.3280
NYCU-NLP0.2973

pan

TeamScore
NYCU-NLP0.5441
SMASH0.5407
AIvengers0.5290
baseline0.4561

spa

TeamScore
SMASH0.5409
NYCU-NLP0.5198
baseline0.5088
PolaFusion0.5065

swa

TeamScore
SMASH0.5840
Aaronbundi0.5830
AIvengers0.5652
baseline0.2205

tel

TeamScore
baseline0.6738
SMASH0.4445
PolaFusion0.4293
Sagarmatha0.4244

tur

TeamScore
baseline0.7693
NYCU-NLP0.5381
happynewyear0.5374
PolaFusion0.5151

urd

TeamScore
NYCU-NLP0.8213
SMASH0.8211
YEZE0.8152
baseline0.5316

zho

TeamScore
NYCU-NLP0.7191
suuii0.7004
SMASH0.6774
baseline0.0000

Visual Summary

These charts provide a quick visual read of the same leaderboard data shown in the tables below.

Top-3 Placement Charts

Stacked bars count first-, second-, and third-place finishes across language-specific leaderboards.

Top 8 teams shown

Subtask 1

Polarization Detection

1st2nd3rd
UTokyo Tsuruoka Lab
12
NYCU-NLP
12
PSK
8
Lingo Research Group
5
SMASH
5
taien
3
yunkuang0329
3
OZemi
2

Subtask 2

Polarization Type Classification

1st2nd3rd
NYCU-NLP
14
SMASH
14
UTokyo Tsuruoka Lab
12
Lingo Research Group
6
CoPol
5
Sagarmatha
3
NASIM LAB
2
PolaFusion
2

Subtask 3

Manifestation Identification

1st2nd3rd
SMASH
16
NYCU-NLP
11
PolaFusion
7
Sagarmatha
4
happynewyear
4
AIvengers
4
OZemi
2
YEZE
2

Baseline Gap Charts

Bars show the difference between each language winner and the POLAR baseline.

Top - baseline

Subtask 1

Largest gaps between the top system and the POLAR baseline.

Top - baseline
tel (Sagarmatha)
+0.261
tur (NYCU-NLP)
+0.137
pol (Lingo Research Group)
+0.119
khm (SMASH)
+0.115
deu (NYCU-NLP)
+0.089
hin (PSK)
+0.086
amh (PSK)
+0.085
rus (UTokyo Tsuruoka Lab)
+0.085

Subtask 2

Largest gaps between the top system and the POLAR baseline.

Top - baseline
tur (CoPol)
+0.306
amh (PolaFusion)
+0.298
hau (NYCU-NLP)
+0.276
mya (NASIM LAB)
+0.270
tel (CoPol)
+0.259
deu (UTokyo Tsuruoka Lab)
+0.212
pol (UTokyo Tsuruoka Lab)
+0.201
eng (UTokyo Tsuruoka Lab)
+0.199

Subtask 3

Largest gaps between the top system and the POLAR baseline.

Top - baseline
zho (NYCU-NLP)
+0.719
nep (NYCU-NLP)
+0.581
hin (SMASH)
+0.536
swa (SMASH)
+0.363
fas (SMASH)
+0.293
urd (NYCU-NLP)
+0.290
arb (NYCU-NLP)
+0.255
ben (SMASH)
+0.194

Best Score Heatmap

Each cell shows the highest official score for a language/subtask pair. Darker cells indicate higher scores.

LanguageSubtask 1Subtask 2Subtask 3
amh0.8000.6700.579
arb0.8490.6700.646
ben0.8630.4220.281
deu0.7610.6200.518
eng0.8250.5320.510
fas0.8350.6440.493
hau0.8340.4800.207
hin0.8240.8070.771
ita0.7300.550-
khm0.7740.7050.437
mya0.8910.747-
nep0.9240.8100.713
ori0.8260.6030.330
pan0.8260.5530.544
pol0.8430.650-
rus0.8300.645-
spa0.8030.6810.541
swa0.8110.5690.584
tel0.9050.5730.445
tur0.8330.7770.538
urd0.8200.7980.821
zho0.9310.8440.719

Important: If your team is missing from the leaderboard, it means the system description paper was not submitted. Please contact the organizers by email if you did submit the system description paper but your name is not displayed.

Need the complete rankings? The full official F1 Macro leaderboards now live on a dedicated Leaderboards page, with search, sorting, and subtask filters.

© 2026 POLAR Shared Task. All rights reserved. Adapted from Editorial Template by HTML5 UP