D2C-Routing: Dimension-to-Composition Evidence Routing for Mixed-Origin AI-Generated Text Detection

Xin Chen; Fuwei Zhang; Yiqi Tong; Wei Guo; Yutian Xiao; Fuzhen Zhuang

cs.CLAug 27, 2026

D2C-Routing: Dimension-to-Composition Evidence Routing for Mixed-Origin AI-Generated Text Detection

Authors: Xin Chen, Fuwei Zhang, Yiqi Tong, Wei Guo, Yutian Xiao, Fuzhen Zhuang

Organizations: School of Artificial Intelligence, Beihang University, Beijing 100191, China

Abstract

AI-generated text detection is commonly framed as a binary document-level judgment about whether a text is human-written or machine-generated. This framing breaks down for mixed-origin writing, where content origin and expression origin may differ. We cast mixed-origin detection as dimension-to-composition source attribution, inferring content origin and expression origin before composing them into four collaboration types. We propose Dimension-to-Composition Routing (D2C-Routing), which routes content-side and expression-side evidence to supervised dimension heads before a learned gated composition layer predicts the final label. On MixD2C, a reconstructed split derived from the HART mixed-origin benchmark, our disclosed D2C-Routing-based detector system reaches 0.8603 four-way Avg TPR@1%FPR, 6.5 points above the same-split RACE-local rerun. Core ablations support the routing design, while error analysis shows that distinguishing AI-content/human-expression from fully AI-generated text remains the hardest boundary. Code is available at https://github.com/bystander563/d2c-routing-artifact.

Figures & tables

Figure 1: Mixed-origin text requires dimension-to-composition detection. We represent attribution through two source dimensions: content origin and expression origin. Their composition yields four labels: HH, HA, AH, and AA. Binary detection collapses mixed-origin cases, while D2C-Routing predicts the two source dimensions before composing the final collaboration type.

Panel A: Labels and targets
Label	Content	Expression	C tgt.	E tgt.
HH	human	human	0	0
HA	human	AI	0	1
AH	AI	human	1	0
AA	AI	AI	1	1
Panel B: Official task collapses

Table 1: Label space. HH/HA/AH/AA compose content origin and expression origin. The two derived binary targets, content origin and expression origin, supervise D2C-Routing; Level-1/2/3 are official HART evaluation collapses, not additional training labels.

Figure 2: D2C-Routing architecture. The figure shows the single-model architecture: the observed document feeds content and expression evidence pathways, supervised source heads estimate the two dimensions, and a learned gate composes them into HH/HA/AH/AA predictions. The separate D2C-base Fusion detector system combines multiple D2C-Routing outputs and is reported only in result tables.

Panel of Table 2 — Table 2: Four-way low-FPR comparison. “Published” rows are representative RACE Table 2 values from a related HART reconstruction rather than exact-split reruns; “MixD2C” rows use the identical local split and evaluator. The dagger denotes our rerun of released RACE code. RACE reports Humanized before LLM-Generated; we remap these to AH and AA. The full published reference set and scalar official-task diagnostics are in Appendix A.1 – A.2 .

Figure 3: Same-split low-FPR behavior on MixD2C. We compare RACE-local, D2C-Routing dual encoder, and D2C-base Fusion on the same MixD2C test split. Our rows mainly improve low-FPR ranking on AA, while RACE-local remains competitive on AH.

Group	Model	Setting	AUROC	Avg TPR@1	AA TPR@1	F1-score	Role
Single	RoBERTa-base	3 seeds	0.9832	0.7666	0.5447	0.8940	Text-only
	Flat concat	3 seeds	0.9851	0.7923	0.6113	0.8987	Flat features
	D2C-Routing	shared, 3 seeds	0.9846	0.8110	0.6841	0.8995	Shared encoder
	D2C-Routing	dual, 3 seeds	0.9858	0.8440	0.7701	0.9054	Best single model
Cost	Text-only 5 $\times$ uniform	623.24M / 5 calls	0.9859	0.8241	0.7385	0.8932	Cost control
	Flat features 5 $\times$ uniform	623.87M / 5 calls	0.9880	0.8443	0.7772	0.9093	Cost control

Table 3: Matched controls and core ablations. Single-model rows use the same encoder family; cost rows match total parameter scale and disclose online forward calls. The D2C cost row reports a separately evaluated AH/AA-calibrated audit system; Table 2 reports the final detector system.

Appendix figures & tables18 assets

Supplementary material from the paper’s appendix.

Appendix

Method	Macro AUROC	HH TPR@1	HA TPR@1	AH TPR@1	AA TPR@1	Avg TPR@1
RoBERTa	0.9222	0.9936	0.6806	0.7092	0.6314	0.7537
RoBERTa-DANN	0.9617	0.9688	0.7503	0.7178	0.4889	0.7314
CoCo	0.9767	0.9968	0.7577	0.7943	0.6393	0.7970
SeqXGPT	0.8987	0.9838	0.1523	0.3168	0.1432	0.3990
DeTeCtive	0.9574	0.9862	0.0000	0.7723	0.0000	0.4396
LF-Motifs	0.9820	0.9668	0.6961	0.7562	0.6701	0.7723

Appendix

Table A2: Full RACE Table 2 reference set, displayed in HH/HA/AH/AA order.

Local run	Macro AUROC	Macro-F1	HH TPR@1	HA TPR@1	AH TPR@1	AA TPR@1	Avg TPR@1
RACE-local, 20ep strict rstdt	0.9849	0.9030	0.9888	0.8463	0.7696	0.5752	0.7950

Appendix

Table A3: Local public-code RACE rerun on the exact MixD2C split.

Detector	L3 Essay	L3 ArXiv	L3 Writing	L3 ALL	L2 Essay	L2 ArXiv	L2 Writing	L2 ALL	L1 Essay	L1 ArXiv	L1 Writing	L1 ALL
RoBERTa(ChatGPT)	0.636	0.796	0.653	0.662 (16%)	0.435	0.687	0.498	0.502 (8%)	0.471	0.955	0.606	0.566 (9%)
RADAR	0.692	0.849	0.647	0.728 (14%)	0.566	0.814	0.630	0.687 (10%)	0.705	0.857	0.700	0.758 (20%)
Log-Perplexity	0.868	0.850	0.811	0.799 (33%)	0.364	0.485	0.438	0.473 (11%)	0.769	0.530	0.625	0.576 (6%)
Log-Rank	0.867	0.874	0.813	0.814 (39%)	0.380	0.460	0.441	0.465 (11%)	0.739	0.542	0.611	0.573 (8%)
LRR	0.835	0.909	0.797	0.840 (50%)	0.560	0.616	0.551	0.573 (25%)	0.616	0.576	0.558	0.568 (19%)
Glimpse	0.929	0.869	0.819	0.849 (58%)	0.754	0.737	0.625	0.676 (30%)	0.878	0.719	0.618	0.688 (22%)

Appendix

Table A4: Original HART English detector references under the original HART protocol.

Detector	Proxy LM / model	L1 AUROC / F1 / TPR@5	L2 AUROC / F1 / TPR@5	L3 AUROC / F1 / TPR@5
SpecDetect	gpt2	0.6859 / 0.8808 / 0.2137	0.5877 / 0.6862 / 0.1744	0.6098 / 0.6259 / 0.2013
SpecDetect	gpt2-xl	0.7129 / 0.8864 / 0.2892	0.6415 / 0.6937 / 0.2694	0.6693 / 0.6333 / 0.2958
Fast-DetectGPT analytic	gpt2-xl scoring/reference	0.7205 / 0.8582 / 0.3688	0.6581 / 0.6676 / 0.3419	0.6883 / 0.6179 / 0.3725
Binoculars proxy	observer=gpt2-xl, performer=gpt2	0.6314 / 0.8571 / 0.2687	0.5588 / 0.6667 / 0.2256	0.5845 / 0.6075 / 0.2357

Appendix

Table A5: Local MixD2C scalar-detector reruns for official Level-1/2/3 collapses.

Model	Source	Level-1	Level-2	Level-3
SpecDetect-gpt2-xl	local scalar rerun	0.7129 / 0.8864 / 0.2892	0.6415 / 0.6937 / 0.2694	0.6693 / 0.6333 / 0.2958
Fast-DetectGPT analytic	local scalar rerun	0.7205 / 0.8582 / 0.3688	0.6581 / 0.6676 / 0.3419	0.6883 / 0.6179 / 0.3725
RACE-local	same-split local rerun	0.9985 / 0.9925 / 0.9967	0.9943 / 0.9641 / 0.9731	0.9841 / 0.9365 / 0.9398
D2C-base Fusion	ours, probability collapse	0.9988 / 0.9925 / 0.9962	0.9959 / 0.9702 / 0.9850	0.9906 / 0.9498 / 0.9685

Appendix

Table A6: Same-split official-task collapses on MixD2C. Each cell reports AUROC / F1 / TPR@5%FPR.

Model	Macro AUROC	Avg TPR@1	Macro-F1	HH TPR@1	HA TPR@1	AH TPR@1	AA TPR@1	Use in paper
D2C-Routing (base no-syntax + AA rank)	0.9845	0.8206	0.9051	0.9888	0.8613	0.7549	0.6777	Routed baseline
Frozen D2C-Routing (last layer)	0.9581	0.5444	0.8165	0.9363	0.2825	0.5735	0.3854	Frozen-probe control
Flat all-features RoBERTa-base	0.9873	0.7982	0.9071	0.9875	0.8138	0.7549	0.6368	Feature-stacking control
D2C-Routing (DistilRoBERTa)	0.9776	0.7414	0.8691	0.9788	0.7425	0.6912	0.5530	Small-backbone control
D2C-Routing (Longformer-base, 512 tokens)	0.9802	0.7639	0.8815	0.9825	0.7413	0.7059	0.6261	Backbone/length control
D2C-Routing (DeBERTa-base no-syntax)	0.9867	0.8158	0.9005	0.9875	0.8338	0.7451	0.6970	Alternative backbone control

Appendix

Table A7: Supplemental architecture controls on MixD2C.

Model	Params	Macro AUROC	Macro-F1	Avg TPR@1	AH TPR@1	AA TPR@1
ModernBERT text-only	149.02M	$0.9850\pm 0.0022$	$0.9089\pm 0.0076$	$0.8032\pm 0.0136$	$0.7484\pm 0.0172$	$0.6562\pm 0.0389$
ModernBERT D2C	149.73M	$0.9851\pm 0.0011$	$0.9026\pm 0.0025$	$0.8073\pm 0.0138$	$0.7435\pm 0.0113$	$0.6796\pm 0.0366$
DeBERTa-base D2C	–	$0.9877\pm 0.0009$	$0.9094\pm 0.0076$	$0.8320\pm 0.0157$	$0.7614\pm 0.0142$	$0.7051\pm 0.0372$

Appendix

Table A8: Three-seed modern-backbone comparisons on MixD2C. Values are mean

\pm

standard deviation.

Routing	Macro AUROC	Macro-F1	Avg TPR@1	AH TPR@1	AA TPR@1
Correct role-aligned	$0.9846\pm 0.0010$	$0.8995\pm 0.0100$	$0.8110\pm 0.0165$	$0.7435\pm 0.0242$	$0.6841\pm 0.0310$
Swapped	$0.9845\pm 0.0033$	$0.8950\pm 0.0157$	$0.8076\pm 0.0254$	$0.7222\pm 0.0396$	$0.6991\pm 0.0326$
Fixed random	$0.9849\pm 0.0005$	$0.9058\pm 0.0057$	$0.8034\pm 0.0109$	$0.7451\pm 0.0147$	$0.6476\pm 0.0591$
Flat concat	$0.9851\pm 0.0019$	$0.8987\pm 0.0079$	$0.7923\pm 0.0072$	$0.7516\pm 0.0057$	$0.6113\pm 0.0499$

Appendix

Table A9: Three-seed comparison of correct, swapped, fixed-random, and flat evidence routing on MixD2C.

Model/system	Macro-F1	Macro AUROC	Avg TPR@1	Interpretation
Text-only dev-selected ensemble	0.9047	0.9883	0.8289	Matched ensemble
Flat-concat dev-selected ensemble	0.8903	0.9863	0.8191	Matched ensemble
Calibrated D2C system	0.9100	0.9877	0.8507	Competitive grouped robustness

Appendix

Table A10: Group-aware MixD2C results with matched detector systems. Paired intervals for D2C versus text-only cross zero for Macro-F1 and Avg TPR@1; only the accuracy difference is significant.

Held-out domain	Model	Macro AUROC	Avg TPR@1	Macro-F1	AA TPR@1
Essay	D2C-Routing	0.9000	0.3663	0.6414	0.1834
Essay	Text-only	0.8894	0.3094	0.6198	0.0000
ArXiv	D2C-Routing	0.9495	0.6760	0.6935	0.8006
ArXiv	Text-only	0.9641	0.7464	0.8051	0.7721
News	D2C-Routing	0.9096	0.3847	0.5555	0.3305
News	Text-only	0.9124	0.4363	0.5827	0.3675

Appendix

Table A11: Leave-one-domain-out HART transfer checks.

Dataset	Model	AUROC	F1	Accuracy	TPR@1	TPR@5
RAID test	Text-only	0.7312	0.5464	0.6493	0.2650	0.3435
RAID test	D2C-Routing (no-external)	0.7210	0.5713	0.6270	0.2010	0.2550
RAID nonnative	Text-only	0.9349	0.8639	0.8736	0.5495	0.7363
RAID nonnative	D2C-Routing (no-external)	0.9144	0.8095	0.8242	0.2198	0.6374

Appendix

Table A12: RAID binary transfer checks.

Setting	Supervision	Result	Claim role
FAIDSet clean test ( Ta et al., 2026 )	Target-supervised adaptation	Accuracy 0.9618; Macro-F1 0.9599	External adaptation, current public split
APT-Eval HH/HA ( Saha and Feizi, 2025 )	Zero-shot expression score	AUROC 0.8993; 293/300 paired ranking	Expression-origin transfer
PAN LLM-DetectAIve HH/HA ( Bevendorff et al., 2025 )	Zero-shot expression score	AUROC 0.9392; TPR@1 0.6491 ( $N=18{,}480$ )	Expression-origin transfer
PAN machine-mediated content contrast	Zero-shot content score	AUROC 0.8012 vs matched text 0.7961 ( $N=20{,}703$ )	Small content-origin transfer
MixSet HART-style reconstruction ( Zhang et al., 2024 )	Target-supervised adaptation	Accuracy 0.8504; Macro-F1 0.7811	External reconstructed four-process task
MixSet direct four-way transfer	Zero-shot, no adaptation	Macro-F1 0.2262; AUROC 0.4372	Negative broad-OOD result

Appendix

Table A14: Protocol-separated robustness evidence. FAIDSet and the HART-style MixSet reconstruction use target supervision; APT/PAN use HART-trained scores without target tuning. MixSet four-way labels are reconstructed from released transformation provenance, not native benchmark labels.

Feature group	Content-origin AUROC	Expression-origin AUROC	Four-way Macro AUROC	Four-way Macro-F1
Entity/coherence	0.8261	0.7277	0.7460	0.4363
Expression features	0.9300	0.8515	0.8777	0.5907
RST discourse	0.8240	0.7250	0.7242	0.4100
Entity + RST	0.8842	0.7747	0.7843	0.4794
All evidence	0.9408	0.8595	0.8781	0.6150

Appendix

Table A15: Direct evidence probe results. The table reports dimension-level and four-way probe performance from pre-extracted evidence features.

Figure 4: Heatmap summary of source-dimension and four-way evidence probes.

Model	AUROC	Avg TPR@1	Macro-F1	AA TPR@1
D2C-Routing (full)	0.9846	0.8110	0.8995	0.6841
No RST	0.9814	0.7639	0.8869	0.5447
No entity/coherence	0.9833	0.7868	0.8969	0.6098
No expression features	0.9802	0.7541	0.8831	0.5805
No routed external features	0.9843	0.7988	0.8948	0.6655

Appendix

Table A16: Feature-removal ablations.

Setting	AUROC	Avg TPR@1	Macro-F1	AA TPR@1
D2C-Routing (full)	0.9846	0.8110	0.8995	0.6841
Remove lexical-connective realization	0.9773	0.7444	0.8860	0.5989
Remove rhythm/POS	0.9835	0.7922	0.8943	0.6287
Remove surface-regularity	0.9828	0.7878	0.8911	0.6177
Lexical-connective realization only	0.9808	0.7534	0.8809	0.5578
Rhythm/POS only	0.9825	0.7567	0.8830	0.5638

Appendix

Table A17: Expression-side subgroup ablations.

Composition	AUROC	Avg TPR@1	Macro-F1	AH TPR@1	AA TPR@1
Learned vector gate	0.9846	0.8110	0.8995	0.7435	0.6841
Fixed average of member models	0.9805	0.7754	0.8880	0.7010	0.6072
Concat without gate	0.9814	0.7942	0.8959	0.7141	0.6846
Hard factorized composition	0.9810	0.7868	0.8971	0.7092	0.6213

Appendix

Table A18: Composition alternatives.

Class	Content acc.	Expression acc.	Correct expr. prob.	Four-way acc.
AH	0.9477	0.6438	0.6374	0.6634
AA	0.9721	0.9988	0.9958	0.9723

Appendix

Table A19: Three-seed branch diagnosis for the two AI-content classes.

AI-Generated Text Detection

AI-generated text detection develops methods to distinguish human-written, machine-generated, and AI-edited text.

See all 171 papers

AEyeDE: An Attention-Based Attribution Framework for AI-Generated Text Detection
Apr 13, 2026Aria Nourbakhsh, Adelaide Danilov, Christoph Schommer +1AI-Generated Text Detection Robustness of Machine-Generated Text Detection
Department of Computer Science, University of Luxembourg
Show, Don't TELL: Explainable AI-Generated Text Detection
May 27, 2026Aldan Creo, Suraj RanganathExplainable Artificial Intelligence AI-Generated Text Detection
School of Computing, Information and Data Sciences University of California, San Diego United States of America
Rethinking AI-Generated Text Detection: A Strong Baseline and the Distribution-Shift Problem That Remains
Jul 4, 2026Zhuoer Shen, Mingyi Wang, Shaofeng Zou +1Distribution Shift Robustness AI-Generated Text Detection
Department of Computer Science, University of California, Santa Barbara · School of ECEE, Arizona State University

Apr 13, 2026J/K move · Enter open · S save

AEyeDE: An Attention-Based Attribution Framework for AI-Generated Text Detection

Aria Nourbakhsh, Adelaide Danilov, Christoph Schommer, Salima Lamsiyah

Department of Computer Science, University of Luxembourg

AI-Generated Text Detection Robustness of Machine-Generated Text Detection

Open paper AskPDF arXiv

Detecting AI-generated text is becoming increasingly challenging as modern language models approach human-level fluency and can evade detectors that rely on surface statistics or likelihood-based signals. We propose AEyeDE, an attribution-driven approach to human-AI authorship detection that leverages model attention as a discriminative signal. Specifically, we extract attention-based attribution matrices for both human- and AI-generated text using a proxy Transformer model with white-box access and train a lightweight Convolutional Neural Network to learn representations from these attribution maps. Across encoder-decoder translation settings, our method consistently outperforms a text-only baseline. In decoder-only settings, it performs strongly in generator-specific detection, remains competitive on standard benchmarks, and shows robustness under cross-dataset transfer and alternative-spelling perturbations. We further show that attention maps exhibit recurring local structures whose relative frequencies differ consistently between human- and AI-generated text across datasets and proxy models. These findings suggest that attention-based attribution maps provide a complementary and interpretable signal for AI-generated text detection. We will make the code publicly available to support future research.

Abstract

AI-Generated Text Detection

Explore similar work

AEyeDE: An Attention-Based Attribution Framework for AI-Generated Text Detection

Show, Don't TELL: Explainable AI-Generated Text Detection

Rethinking AI-Generated Text Detection: A Strong Baseline and the Distribution-Shift Problem That Remains