AI Paper Atlas

Research map / Large language models / Reasoning and chain of thought

Token-Efficient Long-Term Interest Sketching and Internalized Reasoning for LLM-based Recommendation

Zhihao Ding, Jinming Li, Shuai Mu, Jieming Shi

ICLR 2026 · Large language models › Reasoning and chain of thought · approach: Routing, pruning and distillation for reasoning

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes a token-efficient method for LLM-based recommendation that sketches long-term user interests and internalizes reasoning, addressing long noisy histories and the cost of CoT at inference.

Problem

LLMs struggle to reason over long, noisy user histories spanning hundreds of items, and truncation discards signals needed for preference reasoning; explicit CoT reasoning also adds inference cost for recommendation.

Approach

Compresses long-term interest into a compact sketch and internalizes CoT-style preference reasoning so that inference does not require generating long explicit reasoning chains (details not available in the abstract).

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add long-term interest sketch to recent history (ablation)answer-only fine-tuned Qwen3-4B · Amazon MoviesMAE: 0.7775 -> 0.7695Improved
add long-term interest sketch to recent history (ablation)answer-only fine-tuned Qwen3-4B · Amazon BooksMAE: 0.3547 -> 0.3536No clear effect
add raw past interactions instead of sketch (ablation)answer-only fine-tuned Qwen3-4B · Amazon MoviesRMSE: 1.3635 -> 1.3723Hurt
replace sketch with LLM-generated user profile (ablation)answer-only fine-tuned Qwen3-4B · Amazon MoviesRMSE: 1.3635 -> 1.3872Hurt
add GRPO-trained explicit CoT over sketch (ablation)AO-SFT (+Sketch) · Amazon MoviesMAE: 0.7695 -> 0.7577Improved
add hidden alignment to internalize CoT (ablation)GRPO-CoT model · Amazon Bookslatency (s/sample): 2.22 -> 0.013Improved
use hidden alignment instead of CE or KD for internalization (ablation)GRPO-CoT checkpoint · Amazon MoviesMAE: 0.7995 (KD) -> 0.7603 (HA)Improved
add cross-entropy loss to hidden alignment (ablation)HA · Amazon BooksMAE: 0.3503 -> 0.3525Hurt
adapt only KV projections instead of all-linear (ablation)LoRA hidden alignment · Amazon BooksMAE: 0.3505 -> 0.3503No clear effect
use brief item text rather than full descriptions for sketch (ablation)SIREN sketch · Amazon BooksMAE: 0.3536 -> 0.3594Hurt
use distance-aware regression reward instead of binary exact-match (ablation)GRPO rating reward · Amazon MoviesMAE: 0.8094 -> 0.7577Improved
use SIREN versus strongest LLM baselineExp3RT · Amazon BooksMAE: 0.4001 -> 0.3510Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy