- Rating
- 1
- Confidence
- 2
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper focuses on addressing the challenge of personalizing LLMs when handling long-term user interaction history. This work proposes PRISM (Progressive Resolution with Independent Section Mapping), which combines incremental hierarchical compression of episodic into long-term memory and a Map--Reduce decomposition that generates each memory level's independent semantic attribute sections in parallel.
Critical strengths
- PRISM stores user profiles as an incrementally compressed hierarchy in which short-term episodic memories feed into long-term memory.
- The authors provide a hardware-independent evaluation and release a reproducible benchmark for long-horizon personalization.
Critical weaknesses
- The paper is hard to follow. There are no figures at all, not even a method overview. Adding an overview diagram and concrete input/output examples would greatly improve clarity.
- Essential details are missing. For example, the “frontier teacher” is never identified (which model, which version, what decoding settings), even though it is central to the results.
- LongProfile appears to be a simple merge of several existing datasets.
Questions for authors
Please refer to the weaknesses detailed above.
Thank you for your constructive feedback and insightful suggestions! Below, we address each of your points in detail:
Method overview diagram: We added a method overview. Figure 1 shows the full pipeline: raw events are compressed per episode into M1; M2 is generated from M1 only, through K parallel section calls followed by one Reduce call; and a router assembles only the sections a query needs. Figure 2 contrasts the refresh path of the monolithic baseline and PRISM.
Frontier teacher model: The frontier teacher is GPT-5.5 at temperature 0.2, accessed through the OpenAI API with reasoning effort set to low. It generates both the monolithic baseline and the PRISM sections under this configuration.
LongProfile construction: Yes. LongProfile is built from the union of MovieLens-25M, Amazon Reviews 2023 and Yelp, and the Setup subsection and dataset appendix state this explicitly. Its contribution is the construction and the protocol rather than the source data: a long-horizon filter (at least 50 interactions spanning at least two years); balanced stratified cohorts across event-count buckets up to 5,000 interactions; a user-disjoint 60/20/20 split with a temporal last-k holdout; and labels given by held-out raw interactions, scored with IR and error metrics rather than an LLM judge. We have not yet reported downstream ranking results on LongProfile. The construction pipeline and datacard will be released with the code at camera-ready.