- Rating
- 2
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper introduces an algorithm for cost constrained auto-bidding that accounts for the fact that sometimes, for platform/scale reasons, the shading factor cannot be updated continuously over time. The algorithm output a criterion that, when satisfied, trigger an update. This criterion is based on a greedy algorithm that merge time intervals. Experiments on AuctionNet and online on live traffic are reported.
Critical strengths
- The idea of merging greedily using marginal values offline to produce a criterion is interesting
- I found the first half of the article well written
Critical weaknesses
I think the work would benefit from another round of polish.
- see my questions: some passages are hard to parse because some quantities are not defined
- the algorithm is only stated in the appendix. More focus should be put on the algorithm and the exact context, which in my opinion is too vague for now. In particular, section 4.2 is too vague.
For now, the contribution might be too narrow.
- the solution is for a specific problem, does the methodology extends outside of the auto-bidding scope ? (btw does it connect with https://hal.science/hal-05456664v1/document ?)
- strong modelling assumptions
- could we get a more insightful theoretical result?
Questions for authors
- please clarify what the optimal threshold is.
- page 5 , line 232, should read instead?
- line 265 page 5, what is and where is it defined?
- how do you justify the assumptions on the market dynamics?
- what is the context that goes into the transformer?
- can you clarify what BidDreamer is? Is it a contribution? An implementation of something that already existed?
Clarifications on Theoretical Details, Algorithm, and Generalizability
We sincerely thank you for the constructive feedback. We are glad that our idea is appreciated. Below, we provide detailed responses to the questions and concerns.
1. Algorithm and Threshold (Q1, W1)
We provide a more detailed description of the algorithm and online execution in Section 4.2. For the definition of optimal threshold, please refer to response to Reviewer acvK.
2. Theorem 1 (Q2)
The subscript is J. Since the discrepancy in Eq. (8) is evaluated at x, the proof applies the mean value theorem to ρJ. From the common first-order conditions in Eq. (6), together with CI′(x)>0 and CJ′(y)>0, we obtain ρI(x)=CI′(x)GI′(x)=λ∗=CJ′(y)GJ′(y)=ρJ(y). Then, ∥ρI(x)−ρJ(x)∥=∥ρJ(y)−ρJ(x)∥. For x=y, the mean value theorem gives a point ξ∈(minx,y,maxx,y)⊆D such that ρJ(y)−ρJ(x)=ρJ′(ξ)(y−x). Since ∣ρJ′(b)∣≥L>0 throughout D, we have ∣ρI(x)−ρJ(x)∣=∣ρJ′(ξ)∣∥y−x∣≥L∣y−x∣. Therefore,∣x−y∣≤L∣ρI(x)−ρJ(x)∣. Combining this with the spending-preserving candidate bound yields ΔL≤2M∣x−y∣2≤2L2M∣ρI(x)−ρJ(x)∣2,where M=MI+MJ. If x=y, the original optimum is already feasible after merging, so ΔL=0 and the bound holds directly. A slope lower bound on ρI instead would support the symmetric argument.
3. Definition of Ur (Q3)
Ur is introduced before Eq. (9). For the pair (Ir,Jr) selected at merge step r, Ur denotes the merge-loss upper bound: Ur=2Lr2Mr(∣ρ^Ir(boff,r)−ρ^Jr(boff,r)∣+KL,rϵr+ηIr+ηJr)2.
4. Market Dynamics Assumptions (Q4, W2.2)
Our merge-loss analysis requires only general regularity conditions on the response curves, without assuming a specific parametric form or temporal stationarity, allowing θt to vary over time. Increasing cost and decreasing marginal efficiency are economically motivated, while the curvature and slope bounds are local regularity assumptions used to derive the merge-loss theoretically. The exponential parameterization satisfies these properties, and Appendix D provides empirical support. Besides, Appendix F verifies these properties for an alternative Hill parameterization, and Table 5 demonstrates competitive performance.
5. Transformer Context (Q5)
6. BidDreamer (Q6)
BidDreamer serves as the bidding policy in our online experiments. It is not a separate contribution, and our contribution focuses on the scheduling mechanism.
7. Generalizability (W2.1)
The underlying methodology addresses a more general resource-constrained sequential decision problem. Consider a horizon of T decision steps, where each step t is associated with a reward function Rt(u) and a resource consumption function Ct(u), u denoting a continuous action. Given a total resource budget B and a maximum number of K action updates, the problem can be formulated as: P,uII∈PmaxI∈P∑t∈I∑Rt(uI)s.t.I∈P∑t∈I∑Ct(uI)≤B,&∣P∣≤K, where P is a partition of the time horizon, and all steps within each interval I share the same action uI. This formulation is independent of auction-specific mechanisms and can be extended to other scenarios, such as LLM inference resource scheduling, energy management, and dynamic pricing.
Besides, CUVET and our work address different problems. CUVET focuses on treatment assignment by partitioning the user feature space to optimize personalized treatment allocation, whereas our work addresses resource-constrained decision scheduling, aiming to determine when actions should be updated under a limited computational budget.
8. Insight (W2.3)
Identifying and preserving critical decision opportunities is essential for effective decision scheduling. Our analysis suggests that decision updates should be driven by meaningful changes in the Decision Landscape. When the underlying reward-cost trade-offs remain similar, additional updates may be unnecessary. Theoretically, we establish that the performance loss induced by sharing an action across adjacent intervals is upper-bounded by a quadratic function of their marginal-efficiency discrepancy, providing a principled criterion for adaptively identifying which decision boundaries can be safely removed and which should be preserved. Our experimental results further support this theoretical insight.