Omni Models process audio, vision, and text within a unified language backbone, yet often produce different answers for semantically equivalent inputs across modalities. Existing representation-level analyses do not distinguish between information that predicts whether disagreement will occur and the representation whose perturbation affects the model's answer. In this paper, we estimate a low-dimensional Answer Subspace from answer-position representations and use it to separate answer-related components from the remaining representation. Across three Omni Models and two multimodal reasoning tasks, information predictive of cross-modal inconsistency lies mainly outside the Answer Subspace, concentrating in a reproducible two-dimensional representation that is not explained by a fixed modality offset. Matched interventions reveal distinct behavioral roles: perturbing the Answer Subspace substantially changes the model's predictions, whereas perturbing the disagreement representation, like a random subspace of the same dimensionality, leaves them almost unchanged. Our key finding is that cross-modal disagreement can be induced by misaligning modality-specific answer representations: perturbing the Answer Subspace for only one modality induces disagreement, whereas applying the same perturbation to both largely preserves agreement. Thus, the representation that predicts disagreement is distinct from the answer-level representation through which disagreement can be induced, showing that linear decodability does not imply behavioral efficacy.
Figures & tables
Figure 1: Layer-wise probe performance over normalized backbone depth. (a) Answer-probe accuracy shows that the model’s prediction becomes linearly decodable from the answer-position representation in the middle layers. (b) Disagreement-probe accuracy shows that disagreement-predictive information lies primarily outside the answer-related representation, with its orthogonal complement outperforming both the answer-related representation and matched random subspaces.
Figure 2: Disagreement-probe accuracy using the leading one, two, or three right singular vectors of the reshaped disagreement-probe coefficient matrix, selected within each training fold. Performance increases substantially from one to two directions, with little additional improvement from the third. Dashed and dotted lines denote the full hidden-state and Answer Subspace probes, respectively. Appendix B.3 reports per-condition results and bootstrap stability.
Figure 3: Answer decoding under option permutation. Despite substantial variation in model behavior, the decoded answer consistently matches the model’s updated prediction. Squares, diamonds, and circles denote content following, letter retention, and decoded-answer agreement, respectively.
Figure 4: Alignment of the recovered subspaces with the modality-discriminating axis. Despite near-perfect modality separability, both subspaces are nearly orthogonal to this axis. Removing the mean modality offset has negligible effect on disagreement-probe accuracy. Appendix B.6 reports the per-condition values.
Figure 5: Disagreement-probe accuracy after controlling for the audio-condition prediction. Filled and open circles denote answer-controlled and unrestricted accuracy, respectively, and dashed ticks mark the label-permutation null. Appendix B.7 reports the strata and sample sizes.
Figure 6: Paired audio and image representations in the Answer Subspace across six model–task conditions. Gray lines connect the two modality conditions of each item. Agreement pairs remain within the same answer region, whereas disagreement pairs separate across answer regions.
Figure 7: Prediction changes under matched interventions across shift magnitudes. Left and right panels show the external and spatial task families, respectively. Lines denote the mean across three models and shaded regions the range across models. Perturbations of the Answer Subspace produce a strong dose-dependent effect, whereas perturbations of the Disagreement-Predictive Subspace remain comparable to a random two-dimensional subspace drawn from the Answer Subspace complement.
Figure 8: Disagreement induced by symmetric and asymmetric Answer Subspace interventions. Applying the perturbation to one modality produces substantially more disagreement than applying it to both under the same layer, magnitude, and operation.
Appendix figures & tables14 assets
Supplementary material from the paper’s appendix.
Appendix
Model
Family
3 blocks
2 blocks
Overlap
Qwen2.5-Omni
Ext.
0.679
0.674
0.972
Qwen2.5-Omni
Spa.
0.704
0.687
0.985
Gemma-4-E4B
Ext.
0.718
0.712
0.988
Gemma-4-E4B
Spa.
0.681
0.662
0.989
MiniCPM-o 4.5
Ext.
0.723
0.723
0.978
MiniCPM-o 4.5
Spa.
0.639
0.638
0.991
Appendix
Table 1: Effect of removing the difference block. The first two columns are balanced accuracy of the disagreement probe on the full complement; Overlap is the mean principal-angle cosine between the two-dimensional subspaces recovered under the two parameterizations.
Model
Family
Layers
Items
Used
Agree
Disagree
Half-dec. layer
Half-dec. depth
Interv. depth
Qwen2.5-Omni
Ext.
29
1350
1350
1038
312
15
53.600
75.000
Qwen2.5-Omni
Spa.
29
465
465
165
300
19
67.900
75.000
Gemma-4-E4B
Ext.
43
1350
1350
528
822
18
42.900
76.200
Gemma-4-E4B
Spa.
43
465
465
213
252
24
57.100
76.200
MiniCPM-o 4.5
Ext.
37
1350
1350
864
486
17
47.200
75.000
MiniCPM-o 4.5
Spa.
37
465
465
141
324
20
55.600
75.000
Appendix
Table 2: Conditions used throughout. Used is the number of items on which the answer letter could be resolved in both conditions; Agree and Disagree split those items by whether the two modality conditions produced the same letter. Half-dec. layer is the layer at which answer-probe accuracy first reaches the midpoint between chance and its final-layer value, and Half-dec. depth the corresponding normalized depth. Interv. depth is the normalized depth of the layer at which interventions are applied.
Model
Family
Inside
Compl.
Random
PCA
Full
Qwen2.5-Omni
Ext.
0.510
0.704
0.511
0.518
0.705
Qwen2.5-Omni
Spa.
0.516
0.692
0.510
0.517
0.706
Gemma-4-E4B
Ext.
0.586
0.692
0.534
0.623
0.699
Gemma-4-E4B
Spa.
0.645
0.687
0.572
0.631
0.692
MiniCPM-o 4.5
Ext.
0.500
0.725
0.679
0.685
0.718
MiniCPM-o 4.5
Spa.
0.495
0.659
0.516
0.544
0.651
Appendix
Table 3: Balanced accuracy of the disagreement probe at the final layer under five feature sets: coordinates inside the Answer Subspace, inside its orthogonal complement, a rank-matched random readout of that complement (mean of five draws), a rank-matched PCA readout of the complement, and the full hidden state. Chance is 0.5 . The complement matches the full state to within 0.014 in every condition, while the two rank-matched readouts of the same complement fall well short of it, so the result does not follow from the complement’s dimensionality alone.
Model
Family
k=1
k=2
k=3
Random k=2
Bootstrap align.
Min
Null
Qwen2.5-Omni
Ext.
0.593
0.692
0.694
0.515
0.733
0.683
0.033
Qwen2.5-Omni
Spa.
0.666
0.694
0.691
0.512
0.725
0.632
0.031
Gemma-4-E4B
Ext.
0.649
0.670
0.664
0.535
0.713
0.674
0.037
Gemma-4-E4B
Spa.
0.629
0.660
0.674
0.566
0.719
0.670
0.039
MiniCPM-o 4.5
Ext.
0.625
0.721
0.718
0.568
0.724
0.693
0.029
MiniCPM-o 4.5
Spa.
0.611
0.639
0.641
0.507
0.747
0.700
0.028
Appendix
Table 4: Number of retained directions and stability of the recovered plane. The first four columns are balanced accuracy of the disagreement probe on k directions of the complement, and on a random two-dimensional subspace of the same complement. Bootstrap align. is the mean principal-angle cosine between the plane recovered from an item-level bootstrap resample ( 30 resamples) and the plane recovered from the full sample; Min is the smallest such value and Null the 95 th percentile obtained when the resampled directions are replaced by random ones.
Agree
Disagree
Model
Family
Aud.
Img.
Aud.
Img.
Qwen2.5-Omni
Ext.
0.992
0.990
0.978
0.955
Qwen2.5-Omni
Spa.
0.939
0.982
0.970
0.997
Gemma-4-E4B
Ext.
0.987
0.991
0.987
0.996
Gemma-4-E4B
Spa.
0.967
0.981
0.937
0.952
MiniCPM-o 4.5
Ext.
0.983
0.987
0.949
0.969
Appendix
Table 5: Agreement between the answer decoded from the final-layer Answer Subspace and the model’s actual prediction, reported separately for each modality condition and for items where the two conditions agree or disagree. Agreement exceeds 0.93 throughout, including items where the two conditions produced different letters.
Model behavior
Decoded answer agrees with
Model
Family
N
Content
Letter
Other
Model
Model ∣ content
Model ∣ letter
Qwen2.5-Omni
Ext.
2700
0.919
0.028
0.053
0.985
0.991
0.895
Qwen2.5-Omni
Spa.
1320
0.618
0.175
0.207
0.948
0.971
0.935
Gemma-4-E4B
Ext.
2700
0.632
0.184
0.184
0.984
0.989
0.978
Gemma-4-E4B
Spa.
1320
0.366
0.305
0.329
0.967
0.969
0.970
MiniCPM-o 4.5
Ext.
2700
0.880
0.043
0.076
0.985
0.990
0.966
Appendix
Table 6: Option permutation, decomposed. N is the number of (item, modality) pairs. Model behavior is how the model’s own prediction moved: to the letter now holding the original content, staying on the original letter, or to a third letter. The remaining columns are the fraction of pairs on which the answer decoded from the Answer Subspace agrees with the model’s new prediction, overall and restricted to the two behavior groups. The readout follows the model in every group, including when the model stays on the original letter.
Pair distance
Modality
Model
Family
Agree
Disagree
Ratio
Excess/dim
Separability
cos answer
cos disagree
Qwen2.5-Omni
Ext.
0.170
0.225
1.324
7.310
1.000
0.054
0.023
Qwen2.5-Omni
Spa.
0.173
0.218
1.260
3.170
1.000
0.034
0.023
Gemma-4-E4B
Ext.
0.681
0.931
1.367
8.620
1.000
0.065
0.036
Gemma-4-E4B
Spa.
0.841
0.899
1.069
4.750
1.000
0.040
0.060
MiniCPM-o 4.5
Ext.
0.892
1.060
1.188
2.920
1.000
0.046
0.017
Appendix
Table 7: Geometry at the intervention depth. Pair distance is the mean distance between the two modality conditions of the same item inside the Answer Subspace, split by cohort, with their ratio. Excess/dim is how much more the excess separation of disagreeing pairs concentrates inside the Answer Subspace than in its complement, per dimension. Separability is the balanced accuracy of a probe predicting which modality the input arrived in from the answer-position state; the two cos columns are the cosine of the principal angle between the single axis that probe defines and each subspace. Removing the mean modality offset changes disagreement accuracy by at most 0.0031 .
Model
Family
Strata
N
Unrestr.
Matched
Null
Qwen2.5-Omni
Ext.
4
675
0.635
0.670
0.509
Qwen2.5-Omni
Spa.
4
233
0.558
0.726
0.587
Gemma-4-E4B
Ext.
4
675
0.642
0.712
0.528
Gemma-4-E4B
Spa.
2
157
0.648
0.703
0.586
MiniCPM-o 4.5
Ext.
4
675
0.683
0.697
0.534
MiniCPM-o 4.5
Spa.
2
171
0.605
0.711
0.524
Appendix
Table 8: Holding the audio-condition prediction fixed. A fully answer-pair-matched comparison is impossible because a fixed pair of predictions determines the disagreement label, so we stratify on the audio-condition prediction alone and pool the within-stratum results. Strata is the number of audio-condition answers that contain both cohorts, N the number of items retained, and Null the 95 th percentile of a label-permutation null computed under the same stratification. The matched values exceed that null in every condition and are no lower than the unrestricted ones; the unrestricted value for Qwen2.5-Omni on spatial falls below the null, which is why the body claim rests on the matched series.
Model
Family
N
Disagree
Dist. disagree
Dist. agree
Ratio
Qwen2.5-Omni
Ext.
1350
312
2.199
0.418
5.263
Qwen2.5-Omni
Spa.
465
300
2.694
1.117
2.413
Gemma-4-E4B
Ext.
1350
822
2.877
0.573
5.021
Gemma-4-E4B
Spa.
465
252
2.247
0.913
2.460
MiniCPM-o 4.5
Ext.
1350
486
2.113
0.896
2.358
MiniCPM-o 4.5
Spa.
465
324
2.615
1.362
1.921
Appendix
Table 9: Median distance between the two modality conditions of the same item, computed in the full three-dimensional Answer Subspace rather than in the two-dimensional projection shown in Figure 6 , in units of the coordinate standard deviation. Disagree is the number of disagreement items among N ; Dist. disagree and Dist. agree are the median pair distances within each cohort and Ratio their quotient.
Subspace
α
Qwen Ext.
Qwen Spa.
Gemma Ext.
Gemma Spa.
MiniCPM Ext.
MiniCPM Spa.
Answer
2
0.040
0.065
0.020
0.110
0.105
0.085
5
0.090
0.135
0.075
0.275
0.180
0.195
10
0.170
0.250
0.295
0.500
0.740
0.810
20
0.640
0.485
0.860
0.760
0.900
0.995
40
0.980
0.895
0.980
0.825
1.000
0.995
Disagreement
2
0.000
0.005
0.000
0.010
0.005
0.005
Appendix
Table 10: Fraction of predictions changed across intervention magnitudes α , measured in target-subspace standard deviations, over 100 items per condition. Answer Subspace interventions increase monotonically and exceed both controls from α=2 onward.
At intervention depth
Before answer forms
Answer
Disagree
Random
Answer
Disagree
Model
Family
α=10
α=40
α=40
α=40
α=40
α=40
Qwen2.5-Omni
Ext.
28
98
1
0
0
0
Qwen2.5-Omni
Spa.
45
96
6
2
1
2
Gemma-4-E4B
Ext.
36
100
0
1
1
0
Gemma-4-E4B
Spa.
56
89
6
18
2
9
Appendix
Table 11: Number of predictions changed out of 100 items. All arms use the same layer, operation and magnitude within a depth group. Perturbing the Answer Subspace at the intervention depth changes 89 – 100 predictions, and the same perturbation applied before the answer becomes decodable changes 0 – 3 . The Disagreement-Predictive Subspace and the random control are both near zero and show no consistent ordering: the random control exceeds the disagreement subspace in three of the six conditions, most visibly for Gemma-4-E4B on spatial , where a random two-dimensional subspace changes 18 predictions. In that condition the two arms are therefore not distinguishable and the claim rests on the gap against the Answer Subspace rather than on an ordering between the controls.
N
Changed at α=10
Settling depth (%)
Model
Family
Agree
Disagree
Agree
Disagree
Δ
Agree
Disagree
Δ
Qwen2.5-Omni
Ext.
69
31
0.159
0.548
0.389
49.900
57.300
7.400
Qwen2.5-Omni
Spa.
34
66
0.471
0.439
-0.031
59.600
59.600
0.000
Gemma-4-E4B
Ext.
59
41
0.390
0.317
-0.073
50.800
49.900
-0.900
Gemma-4-E4B
Spa.
47
53
0.426
0.679
0.254
58.200
62.900
4.700
MiniCPM-o 4.5
Ext.
70
30
0.714
0.833
0.119
48.000
52.900
4.900
Appendix
Table 12: Answer Subspace intervention split by cohort, at the smaller magnitude. Disagreement items change more often in four of the six conditions, and the sign of that difference agrees with the difference in settling depth — the normalized depth from which the decoded answer stops changing — in every condition ( ρ=0.94 , p=0.005 ). At α=40 both cohorts are near ceiling and the difference disappears. Cohort sizes are unequal because the split follows the model’s own behavior, so these comparisons are the least powered in the paper and we do not build any claim in the body on them.
Model
Family
Disagree ( k=2 )
Random ( k=2 )
Random ( k=2 , b)
Answer ( k=3 )
Random ( k=3 )
Qwen2.5-Omni
Ext.
-0.002
+0.002
+0.008
-0.103
-0.033
Qwen2.5-Omni
Spa.
+0.002
-0.000
-0.000
-0.041
-0.009
Gemma-4-E4B
Ext.
+0.008
+0.043
-0.027
-0.054
+0.006
Gemma-4-E4B
Spa.
+0.003
-0.004
-0.007
+0.039
-0.002
MiniCPM-o 4.5
Ext.
-0.015
+0.006
-0.013
-0.416
-0.019
MiniCPM-o 4.5
Spa.
-0.005
-0.001
-0.011
-0.027
-0.018
Appendix
Table 13: Change in the top-two answer margin after removing each subspace, over 240 (item, modality) pairs. Values in bold differ from zero at the 95% level under a paired bootstrap. Two independent random draws at k=2 are shown to indicate the spread of the control itself.
Asymmetric
Symmetric control
Model
Family
N
2
5
10
20
40
2
5
10
20
40
Qwen2.5-Omni
Ext.
60
0.000
0.000
0.017
0.500
0.983
0.033
0.083
0.183
0.300
0.050
Qwen2.5-Omni
Spa.
60
0.033
0.033
0.100
0.367
0.833
0.200
0.300
0.550
0.600
0.233
Gemma-4-E4B
Ext.
60
0.000
0.000
0.200
0.817
0.950
0.033
0.133
0.383
0.183
0.083
Gemma-4-E4B
Spa.
60
0.000
0.217
0.400
0.700
0.767
0.067
0.283
0.433
0.550
0.250
MiniCPM-o 4.5
Ext.
60
0.017
0.033
0.717
0.950
1.000
0.083
0.217
0.233
0.200
0.100
Appendix
Table 14: Fraction of originally agreeing items on which disagreement appears, by intervention magnitude α in units of the coordinate standard deviation. Asymmetric perturbs the Answer Subspace under one modality only; the symmetric control applies the identical shift to both. N is the number of originally agreeing items evaluated. The asymmetric arm exceeds the control in all six conditions at α=40 (by at least a factor of 3.07 ), in five of six at α=20 , and in none at α≤5 .
Multimodal model interpretability studies how AI systems combine and use images, text, audio, video, and other signals, and how to explain their decisions.