arXiv Computation and Language

Listening to the Wise Few: Query-Key Alignment Unlocks Latent Correct Answers in Large Language Models

arXiv AI
Oct 12

First-Order In-Context Shapley: Scalable Data Auditing for LLM Alignment and Evaluation

The paper introduces First-Order In-Context Shapley, an inference‑only method that evaluates data utility for large language model alignment by computing the first‑order term of an in‑context cooperative game over semantic k‑NN neighborhoods in O(k) forward passes. It projects log‑likelihood shifts onto a directed influence graph, yielding three signals—Exemplifying Advantage, Exemplified Advantage, and One‑of‑a‑Kind Score—to identify high‑utility demonstrations and rare prompts. The approach reduces compute by 19× compared to Monte‑Carlo Shapley, filters 1–5% of records, and improves error detection and reward accuracy in DPO training and benchmark auditing. whyItMatters":"By enabling scalable, inference‑only data auditing, the method addresses label noise and safety inversions that hinder post‑training alignment of large language models, leading to measurable improvements in model reliability and reward accuracy."

By Yunting Song, Matthew Watson, Moustafa Farid Alzantot, Peter Grabowski, Jun Qin
arXiv Computation and Language
Oct 26

Semantically Rich, Metrically Affect-Blind: A Geometric Blind Spot in Modern Text Embedders

The paper uncovers a systematic mismatch in modern text embeddings: while sentiment can be extracted with high accuracy from frozen encoders, the embeddings’ cosine similarity metric barely reflects polarity differences. Across eleven models from 2021–2026, a linear probe achieves up to 91.3% accuracy, yet cosine similarity margins are only 0.003–0.064, leading to poor retrieval performance (e.g., Qwen3-Embedding-4B’s 64.3 MAP). Introducing a lightweight supervised‑contrastive head (Sentiment‑Injected Sentence Embeddings) dramatically improves probe accuracy, retrieval MAP, and polarity margins, but at the cost of downstream semantic similarity. "whyItMatters":"The findings reveal that widely used retrieval and clustering systems may ignore valuable sentiment information encoded in embeddings, impacting their effectiveness."

By Mahdi Naser Moghadasi
arXiv Machine Learning
Oct 23

Joint or Per-Option? Compact Encoders for Closed-Set Decisions from 17 Million to 1 Billion Parameters

The paper investigates how compact an encoder can be for closed‑set decision tasks in language‑model pipelines, comparing joint versus per‑option scoring. In a 32‑million‑parameter study, per‑option scoring outperforms joint scoring by 5.8 chance‑corrected skill points on a 37‑benchmark suite, and two external benchmarks corroborate this advantage. Across six released joint encoders ranging from 17 million to 1 billion parameters, the 68‑ and 150‑million‑parameter models surpass a 435‑million‑parameter DeBERTa‑v3 zero‑shot classifier on 209 overlapping tasks and match it on 145 unseen tasks, while the smallest model processes a median request in 14 ms versus 728 ms for the classifier.

By Tardelli Ronan Coelho Stekel
arXiv Machine Learning
Nov 2

From the Zoo to the BULLPEN: Sample-Efficient LLM Profiling via Multi-Channel Model Embeddings

The paper introduces BULLPEN, a framework that builds compact embeddings of large language models (LLMs) by jointly analyzing model answers, correctness, and question text. BULLPEN expands on prior work by adding 20 new benchmarks focused on behavior, style, and provenance, and demonstrates that embeddings derived from only 400 questions retain most of the predictive power of full‑pool methods. Across 260 open‑weight models, BULLPEN’s answer‑text encoder achieves higher accuracy on new tasks and preserves over 91% of full‑pool skill with minimal inference cost.

By Russell Ro, Qiming Cui, Reza Abbasi-Asl
arXiv AI
3d ago

Small Heads, Strong Decisions: A Controlled Study of Typed-Decision Models

The paper reports a controlled study of typed‑decision models, re‑implementing five different readout heads on a single backbone (Qwen3‑0.6B) with a unified training protocol. Results show that head complexity has little impact on accuracy, while the backbone size is the dominant factor for performance gains. The study also finds that batching all questions in one prompt yields comparable accuracy with faster inference and training.

By Ritvik Rastogi
arXiv Machine Learning
Oct 8

Latent Performance Profiling of Large Language Models

arXiv:2605.30018v3 Announce Type: replace-cross Abstract: Large language models (LLMs) frequently achieve impressive scores on standardized benchmarks, yet accuracy alone offers a limited view of the...

By Tanmoy Chakraborty, Ayan Sengupta, Suparna Bhattacharya, Partha Pratim Chakrabarti, Amlan Chakrabarti, Supratik Chakraborty, Partha Pratim Das, Lipika Dey, Richa Singh, Mayank Vatsa