# Compile the Table: Query-Calibrated Operator Compression for Tabular In-Context Learning

arXiv: https://arxiv.org/abs/2610.11784 · HTML: https://mlsanity.com/paper/2610.11784/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 36 references that are in the map.
- First posted on arXiv 8 Oct 2026.

## Builds on (most cited first, 18 of 18)

- [PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling](https://mlsanity.com/paper/2406.02069/) (2024)
- [Dataset Distillation](https://mlsanity.com/paper/1811.10959/) (2018)
- [Long Context Compression with Activation Beacon](https://mlsanity.com/paper/2401.03462/) (2024)
- [Model Tells You Where to Merge: Adaptive KV Cache Merging for LLMs on Long-Context Tasks](https://mlsanity.com/paper/2407.08454/) (2024)
- [TuneTables: Context Optimization for Scalable Prior-Data Fitted Networks](https://mlsanity.com/paper/2402.11137/) (2024)
- [Mixture of In-Context Prompters for Tabular PFNs](https://mlsanity.com/paper/2405.16156/) (2024)
- [In-Context Data Distillation with TabPFN](https://mlsanity.com/paper/2402.06971/) (2024)
- [Scaling TabPFN: Sketching and Feature Selection for Tabular Prior-Data Fitted Networks](https://mlsanity.com/paper/2311.10609/) (2023)
- [Latent Context Compilation: Distilling Long Context into Compact Portable Memory](https://mlsanity.com/paper/2602.21221/) (2026)
- [Chunked TabPFN: Exact Training-Free In-Context Learning for Long-Context Tabular Data](https://mlsanity.com/paper/2509.00326/) (2025)
- [No Mean Feat: Simple, Strong Baselines for Context Compression](https://mlsanity.com/paper/2510.20797/) (2025)
- [GRKV: Global Regression for Training-Free KV Cache Compression in Long-Context LLMs](https://mlsanity.com/paper/2605.31105/) (2026)
- [KVSculpt: KV Cache Compression as Distillation](https://mlsanity.com/paper/2603.27819/) (2026)
- [CRUMB: Efficient Prior Fitted Network Inference via Distributionally Matched Context Batching](https://mlsanity.com/paper/2606.11473/) (2026)
- [Every Cache Entry Earns Its Place: Global Allocation of Resolution and Coverage for KV Cache Compression](https://mlsanity.com/paper/2608.07001/) (2026)
- [LUCoS: Latent Unsupervised Context Selection for Tabular Foundation Models](https://mlsanity.com/paper/2605.27254/) (2026)
- [Balanced Adaptive Prototype Selection for Scalable TabPFN Inference on Large-Scale Tabular Data](https://mlsanity.com/paper/2608.12989/) (2026)
- [ARASH: Adaptive Retrieval And Shot Selection for Tabular Prediction](https://mlsanity.com/paper/2608.17856/) (2026)

## Abstract (arXiv)

Tabular in-context learning (ICL) has emerged as a training-free and accurate paradigm for tabular prediction, but current approaches to compressing its in-context examples face an accuracy-throughput tradeoff: fixed subsets can sacrifice accuracy, while query-specific retrieval limits cache reuse and batching across queries, reducing throughput. We propose QCOC (Query-Calibrated Operator Compression), which exploits the exchangeability and repeated use of in-context examples by compiling their full KV cache once into compact memory shared across subsequent queries. Instead of retaining raw examples, QCOC clusters their states into joint-KV prototypes, preserves per-cluster multiplicities and the original example count, and calibrates prototype values against attention query vectors produced by the in-context examples through an anchored closed-form solution. Prototype compression drives the speedup, while value fitting helps preserve accuracy. On 64 held-out OpenML-CC18 datasets, QCOC achieves the highest mean accuracy among the compared compression and retrieval methods at both retained counts. Across 12 configurations on seven long tables, it ranks first among compressed methods in ten and averages 0.23 percentage points below full context. Compressing 8,192 in-context examples to 512 memory slots yields a 10.5x cache compression ratio; excluding one-time compilation, in a single-core CPU online-serving comparison over 1,000 queries, QCOC is up to 508x faster than dynamic retrieval baselines and 1.98x faster than full-context inference. These results show that QCOC enables compact-memory reuse and efficient inference across queries while retaining accuracy close to full context.