- Rating
- 2
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper proposes TDE to turn attribution from a frozen reference model into a representation-learning signal. For an observed training target, it defines a score over the generated target tokens and traces contributions from the target to the history through feature gradients and decoder--encoder cross-attention. Positive attributions become weights over historical tokens. Two independently perturbed versions of these weights produce views for contrastive training. At inference, the recommender generates from history alone. On four Amazon datasets, the paper reports gains for two generative backbones and compares TDE with heuristic augmentation, generic contrastive learning, same-target pairing, and uniform weighting. It also studies attribution choices, evidence views, and transfer across architectures.
Critical strengths
-
The target-conditioned supervision is well motivated. The generation loss tells the model which item to predict, but not which past interactions matter for that prediction; different targets may draw on different parts of the same history. TDE turns that distinction into a training signal. I see this as the paper's strongest motivation.
-
The paper gives a concrete path from attribution to supervision. It defines a score over the generated target tokens, traces contributions through decoder--encoder cross-attention, and uses two evidence-grounded views to train representations rather than stopping at a post-hoc explanation. Comparisons with heuristic augmentation, generic contrastive learning, same-target pairing, and uniform weights make the value of this design more credible.
Critical weaknesses
-
The paper does not independently test attribution faithfulness. Target-dependent weights and better recommendation scores do not show that high-weight interactions actually support the target prediction. Since the reference and student start from the same backbone, the auxiliary loss may also reinforce mistaken reference-model preferences. This leaves the central “decision evidence” interpretation unconvincing.
-
The advance over prior work is unclear. Explanation-guided contrastive learning for sequential recommendation has been studied, for example in EC4SRec (CIKM 2022; arXiv:2209.01347). Adapting attribution to generated target tokens is a concrete contribution, but the paper does not establish what this adaptation adds over earlier explanation-guided methods. This limits my assessment of its novelty.
Questions for authors
-
Implementation clarification: Did the experiments use the sum normalization in Eq. (8) or the Softmax in Algorithm 1? If Softmax was used, zero-attribution tokens still receive positive weight. How should the claim of retaining only positive evidence be understood?
-
Attribution validity: Beyond downstream ranking metrics, how do you establish that high-weight history items support the reference model's target prediction? Could you compare target-score changes after perturbing high- and low-attribution items, and test whether the gains survive mismatching weights with training targets?
-
Negative evidence: What is the intuition for discarding negative attribution? Is the signed ablation in Fig. 3d comparable in normalization and view construction? Could negative attribution carry useful opposing evidence?
-
Novelty: Relative to explanation-guided contrastive sequential recommendation such as EC4SRec (CIKM 2022; arXiv:2209.01347), what capability or gain specifically comes from target-token attribution and the two evidence-grounded views?
-
Scope: TDE hurts BERT4Rec on Baby and Sports in Table 4. Under which architectures or training conditions should it help? Although inference is unchanged, does the training cost in Table 10 limit practical use?
4 comments
W1/Q2 Part II: Target–Evidence Mismatch; W2/Q4: Novelty and Comparison with EC4SRec
W1/Q2 Part II: Target–Evidence Mismatch
We conduct a WrongTarget ablation, following the target-substitution setting in Fig. 1 of our paper. Specifically, we keep the same historical sequence and original near-future training target, but replace the attribution evidence with that extracted for a temporally later target from the same user's interaction sequence. WrongTarget evidence is constructed exclusively from training interactions. The recommendation objective remains unchanged, while the attribution evidence is conditioned on a different target. The dual-view construction and contrastive learning objective remain unchanged.
- Results of target–evidence mismatch on Amazon Office.
| Method | R@5 | R@10 | N@5 | N@10 |
|---|---|---|---|---|
| ActionPiece | 0.0525 | 0.0835 | 0.0343 | 0.0446 |
| TDE-WrongTarget | 0.0528 | 0.0869 | 0.0343 | 0.0452 |
| TDE | 0.0577 | 0.0900 | 0.0376 | 0.0479 |
- Results of target–evidence mismatch on Amazon Tools.
| Method | R@5 | R@10 | N@5 | N@10 |
|---|---|---|---|---|
| ActionPiece | 0.0268 | 0.0444 | 0.0171 | 0.0229 |
| TDE-WrongTarget | 0.0305 | 0.0490 | 0.0202 | 0.0262 |
| TDE | 0.0324 | 0.0499 | 0.0216 | 0.0272 |
- Observations.
- TDE consistently outperforms TDE-WrongTarget. Mismatching the target evidence reduces the gains.
- Together with Fig. 1 in our paper, these results provide complementary evidence. Fig. 1 demonstrates that attribution changes with the prediction target, while this WrongTarget ablation demonstrates that correctly aligning this attribution with the training target is important for effective representation supervision.
W2/Q4: Novelty and Comparison with EC4SRec
We implement EC4SRec-style learning under the same ActionPiece backbone and training budget. Following EC4SRec, this baseline uses importance-guided cropping, masking, and reordering to construct augmented histories, together with its contrastive objective EC4SRec(SSL). Results on Amazon Office and Tools are as follows.
- Comparing TDE with EC4SRec-style learning on Office.
| Method | R@5 | R@10 | N@5 | N@10 |
|---|---|---|---|---|
| ActionPiece | 0.0525 | 0.0835 | 0.0343 | 0.0446 |
| AP+CL4SRec | 0.0522 | 0.0862 | 0.0327 | 0.0437 |
| AP+EC4SRec(SSL) | 0.0495 | 0.0854 | 0.0332 | 0.0447 |
| TDE | 0.0577 | 0.0900 | 0.0376 | 0.0479 |
| Improvement (%) | +16.57 | +5.39 | +13.25 | +7.16 |
- Comparing TDE with EC4SRec-style learning on Tools.
| Method | R@5 | R@10 | N@5 | N@10 |
|---|---|---|---|---|
| ActionPiece | 0.0268 | 0.0444 | 0.0171 | 0.0229 |
| AP+CL4SRec | 0.0288 | 0.0463 | 0.0197 | 0.0253 |
| AP+EC4SRec(SSL) | 0.0302 | 0.0474 | 0.0203 | 0.0258 |
| TDE | 0.0324 | 0.0499 | 0.0216 | 0.0272 |
| Improvement (%) | +7.28 | +5.27 | +6.40 | +5.43 |
- Observations.
- TDE consistently outperforms AP+EC4SRec across all eight metrics on both datasets.
- TDE is effective beyond explanation-guided discrete sequence augmentation.
- Novelty of TDE comparing with EC4SRec.
- TDE explicitly attributes the target token generation in autoregressive generative recommendation to historical interactions.
- TDE extracts fine-grained, token-level, target-conditioned decision evidence by tracing the decoder-to-encoder cross-attention pathway.
- TDE transforms decision explanations into representation-level supervision by constructing two evidence-weighted contrastive views, beyond directly modifying historical sequences through explanation-guided augmentation.
Q5: architectures; Training Cost
Q5 Part I: Under which architectures should TDE help?
In fact, we have explicitly investigated this issue in Section 5.5 (RQ5), Appendix E, and Table 18.
Empirical evidence across architectures
Table 18 (Appendix E) summarizes the transferability of TDE across homogeneous and heterogeneous architectures.
Table: Average improvements are computed over four datasets and four evaluation metrics, as reported in Table 18.
| Evidence Transfer | Architecture | Decision-Pathway Alignment | Objective Alignment | Avg. Improvement |
|---|---|---|---|---|
| ActionPiece → ActionPiece | Encoder–Decoder | Strong | Strong | +10.7% |
| GenPAS → GenPAS | Encoder–Decoder | Strong | Strong | +11.2% |
| SASRec → SASRec | Decoder-only | Strong | Strong | +31.2% |
| BERT4Rec → BERT4Rec | Encoder-only | Weak / Entangled | Weak / Potential conflict | -2.9% |
| ActionPiece → SASRec | Encoder–Decoder → Decoder-only | Strong | Strong | +34.3% |
TDE benefits both encoder–decoder (ActionPiece, GenPAS) and decoder-only (SASRec) architectures, and even heterogeneous ActionPiece → SASRec.
When should TDE help? A twofold alignment principle
As discussed in Section 5.5 and Appendix E, the effective transfer depends on two complementary conditions.
- Decision-evidence alignment: The reference model should expose a meaningful target-to-history decision pathway, allowing attribution to identify historical evidence supporting the target. ActionPiece provides decoder–encoder cross-attention, while SASRec provides causal self-attention at the prediction position.
- Representation-objective alignment: Evidence-guided contrastive supervision should be compatible with the trainable backbone's recommendation objective. ActionPiece and SASRec both optimize future-target prediction, making their representation objectives naturally aligned with TDE's target-conditioned evidence supervision.
Why does BERT4Rec behave differently? Appendix E provides explanations
- BERT4Rec's bidirectional self-attention produces globally contextualized historical representations, making individual historical contributions less directly attributable (attribution entanglement).
- Its Cloze-style masked-item reconstruction objective differs from the future-target prediction objective underlying TDE, potentially creating conflicting representational preferences.
- Consistent with this weaker alignment, TDE improves BERT4Rec on Office and Tools for NDCG metrics but degrades performance on Baby and Sports (Table 4). These observations are consistent with our proposed alignment hypothesis.
Q5 Part II: Training Cost and Practicality
We report the complete training time (seconds), including backbone pretraining (Stage I) and continuation training (Stage II).
Note, Shared Stage I Pretraining:
- Stage I is performed only once for each backbone and dataset, rather than repeated separately for the baseline and TDE. Specifically, we select the pretrained checkpoint with the best validation performance and use this same checkpoint to initialize both AP-continued and TDE-AP for their respective Stage II continuation training.
- The same procedure applies to GenPAS-continued and TDE-GenPAS.
| Dataset | Method | Stage I (s) | Stage II (s) | Total (s) | Change |
|---|---|---|---|---|---|
| Office | AP-continued | 3,455 | 5,030 | 8,485 | — |
| Office | TDE-AP | 3,455 | 7,676 | 11,131 | +31.18% |
| Office | GenPAS-continued | 4,390 | 12,600 | 16,990 | — |
| Office | TDE-GenPAS | 4,390 | 7,160 | 11,550 | -32.02% |
| Tools | AP-continued | 12,462 | 12,462 | 24,924 | — |
| Tools | TDE-AP | 12,462 | 19,838 | 32,300 | +29.59% |
| Tools | GenPAS-continued | 12,480 | 17,430 | 29,910 | — |
| Tools | TDE-GenPAS | 12,480 | 7,140 | 19,620 | -34.40% |
- ActionPiece bacbone: Compared with AP-continued, TDE-AP increases total training time on the two datasests.
- GenPAS bacbone: Compared with GenPAS-continued, TDE-GenPAS reduces the total training time on the two datasests. Its Stage II continuation time is reduced.
- Inference efficiency: TDE uses the frozen reference model and evidence-guided contrastive learning exclusively during training. At inference, these components are removed, and the trained backbone performs standard recommendation without additional computational overhead.
Rebuttal follow-up.
Thank you again for your valuable time and comments on our paper. As the discussion period ends soon, we would like to know whether our responses and revised PDF address your concerns. If you have any further questions or suggestions, we would be very happy to discuss them. Many thanks!
Q1/Q3: Normalization and Negative; W1/Q2: Faithfulness
Q1/Q3: Attribution Normalization and Negative Evidence
We investigate attribution normalization and evidence polarity on Amazon Tools.
W1/Q2 Part I: Attribution Faithfulness
We conduct two complementary intervention experiments on Amazon Tools, using the same frozen reference model:
We compare four strategies: High (highest attribution), Low (lowest positive attribution), Random, and Recent (most recent interactions). All interventions are performed at the item level, followed by re-tokenization.
ΔlogP denotes target log-probability drops relative to the full history.
High-attribution items are more influential
Table: Deletion faithfulness: higher values indicate greater prediction degradation.
High-attribution items preserve predictive information
Table: Evidence sufficiency: lower is better.