AI & Machine Learning · Computer Science & AI
Computation & Language / NLP
The newest papers in Computation & Language / NLP that ReadSearch has read, newest first. Each links to its own page, with its authors, date, where it was published and the original.
The newest 5,000 papers, from 17 September 2026 to 5 November 2026: 5,000 arXiv preprints.
Get the few papers worth your time in Computation & Language / NLP every morning.
Follow Computation & Language / NLP and ReadSearch reads its new papers each day, then brings you the few worth reading, each with its key finding in a sentence.
Newest papers, page 8
On the Opportunities and Challenges of Prefill-to-Autoregressive Cross-Model Transfer
Prefill-to-Autoregressive Cross-model Transfer (PACT), a form of latent communication between language models, combines a larger model's prompt representations with a smaller model's efficient…BlockLock: Reproducible LLM Inference with Fixed Batch Geometry
Reproducible LLM inference requires that a request produces the same output regardless of the other requests in the batch.LowRAM: Low-Rank Keys and Amortised Values for Composable KV Cache Compression at Test-Time
Autoregressive decoding relies on the KV cache, whose memory footprint grows linearly with context length and caps batch size and throughput.Follow the Policy, Fail the Benchmark: Auditing and Revising Policy-Based Decision Benchmarks
Policy-based decision benchmarks evaluate whether systems apply written rules to financial and operational cases.Closed-Loop Agent Evaluation: Layered Measurement and Self-Calibration
Once an agent is live, teams want to know whether it is doing its job, not just whether it throws errors.Reclassification Is All You Missed: Surfacing Variant Reinterpretation Need in Online Health Communities at Population Scale
Genetic variants are routinely reclassified in ClinVar after a patient is tested, sometimes from uncertain to pathogenic, yet patient recontact is inconsistent, with no clear duty and responsibility…Automated Feedback to Peer Tutors Increases Their Talk Moves and Learner Reasoning but Not the Balance of Talk
Automated feedback on classroom discourse improves instruction in professional and online teaching contexts, and recent work suggests that delivering similar feedback to learners may strengthen…Consistent yet Contextual: Learning Author-Level Mixtures of Style for Authorship Attribution
Approaches to authorship verification or attribution typically model the style of a document, asking whether two documents were likely written by the same author.What Patients Say They Can't Access: Mining Cancer Screening Barriers from Online Communities
Cancer screening disparities persist across the U.S., yet the patient-reported barriers driving low uptake remain poorly characterized at scale.Adapting LLM Judges to Observed Human Annotation Policy through Soft Prompt Tuning
Large Language Models (LLMs) are increasingly employed as automated judges for evaluating chatbot responses.Matriarchy Classics: A Computational Approach to Swap Only Gendered Words in Classic Novels
This paper proposes a trustworthy computational method to gender-swap public domain novels, with programmatic guarantees that only gendered words are inverted and all other words are left unchanged.τ-bench-LILT: Evaluating Conversational Agents on Realistic Multilingual Tool-Use Tasks
Frontier evaluations of tool-using conversational agents are conducted almost exclusively in English, leaving a critical gap for organizations deploying agents in other languages and regions.Compositional Unalignment: Auditing Safety in Large Vision-Language Models
Large vision-language models (LVLMs) are vulnerable to compositional harm: an image and a prompt that are benign in isolation become harmful through the model's cross-modal reasoning.SAFE-Teen: Evaluating Chatbot Safety Across Adolescent Mental Health Conversations
Adolescents are increasingly turning to AI chatbots for mental health support, advice, and companionship, raising safety concerns.Source-balanced Capped Fisher: Bounded Evidence Aggregation for User Inference in Language Models
User inference asks whether a source contributed documents to a language-model update, even when the auditor observes only fresh documents from that source.Recursive Sparse Attention: Million-Token Length Extrapolation at Logarithmic Cost
At context lengths of about 1M tokens, dense attention is costly in prefill and decoding, and training typically relies on long documents, which become scarcer as the target length increases.Archetypal contexts in the neurons of language models: an architecture for intelligence?
We validate our previous hypothesis that a language model holds relational patterns apart from the subject domains that instantiate them (Nordfors, 2026).How Do AI Agents Waste Your Money?
AI agents may consume massive tokens across a wide range of tasks, but how many of those tokens are really necessary?$D^2$-OPD: Separating Direction and Distance in Extrapolative On-Policy Distillation
On-policy distillation (OPD) trains a student model to match a teacher's next-token distributions on student-generated prefixes.Swapping the Judge Changes the Ranking: A Preregistered Study of Cross-Provider Replication in LLM-Judged System Comparisons
Practitioners who compare systems with an LLM judge and have no human labels often check a conclusion by repeating it with a judge from another provider.