Exploring the Design Space of Permutation Diffusion LLMs
Permutation generation with language models requires both text-conditioned ranking and a global one-to-one assignment between candidates and output positions.
arXiv - cs.CL (NLP)2 min read
AI News
Stay updated with the latest developments in Artificial Intelligence
APRiL: An Open-ended Benchmark for Chinese Social Awareness in Large Language Models
Large language models are increasingly used in social interactions, where appropriate responses require recognizing implicit cues, inferring mental states, identifying conflicts, and considering potential consequences.
arXiv - cs.CL (NLP)2 min readVertical Sliced Wasserstein Distances for Meta-Measures
Many modern learning problems require comparing probability distributions whose individual observations are themselves probability measures, naturally giving rise to meta-measures in $\mathcal{P}(\mathcal{P}(\Omega))$.
arXiv cs.LG2 min readMeddeX: A Retrieval-Augmented LLaMA-3 System with Gemini Vision for Evidence-Backed Medical Report Analysis
Modern healthcare generates increasingly complex medical reports that challenge clinical professionals in terms of time and interpretive burden. Manual review processes are slow, error-prone, and difficult to scale.
arXiv - cs.CL (NLP)2 min readWolof-Adapted Industry-Ready Open-Weight Language Models
Wolof is spoken by 23M people, mostly in Senegal, but remains weakly supported by open-weight language models, while API models raise cost, governance, and sovereignty constraints for local deployments.
arXiv - cs.CL (NLP)2 min readFlowBudget: Training-Free Sampling-Budget Estimation for Flow Matching Models
Sampling from continuous flow-based generators such as flow matching models requires numerical ODE or SDE integration, making inference cost depend on the number of function evaluations (NFEs).
arXiv cs.LG2 min readCan LLMs Anticipate Their Cruxes?
To use language models to help us make decisions we would like to understand the rationale behind the language model's outputs. In principle, the language model provides its rationale as part of its text output. However, the sheer volume of text makes this hard to use in practice.
arXiv - cs.CL (NLP)2 min readAbridgeKV: Compressing Long Reasoning into the KV Cache of Learned Pause Tokens
Long chain-of-thought decoding is memory-bound: the KV cache of the generated tokens grows linearly with the test-time compute being spent. Evicting cache entries does not fix this, because a model never trained on a cache with entries removed generates longer, often up to the length limit.
arXiv - cs.CL (NLP)2 min readReference-Guided Turn-Taking Style Transfer for Full-Duplex Spoken Language Models
Full-duplex spoken language models decide at every frame whether to take the floor, defining a conversational personality -- yet each set of weights encodes exactly one, which human speakers bracket rather than share.
arXiv - cs.CL (NLP)2 min readContext Injection and Long-Horizon Reliability: Separating Duplicate Recognition, Local Processing, and Counter Tracking
Long-horizon failures of language models are usually measured as end-task success or as a horizon length. We instead score, step by step, three components of a 400-step ledger task in which every step is easy in isolation and the model's own answers are re-injected into its context: recognising a…
arXiv - cs.CL (NLP)2 min readTurboMamba: Faster Inference, Faster Training, and Better Prediction with Lexical Memory
We introduce TurboMamba, a language-model architecture for faster inference, faster training, and better token prediction. A global Mamba-3 model runs once per completed token block. A small Transformer uses this global context and earlier tokens within the block to predict the next token.
arXiv cs.LG2 min readPRISM: Transfer and Limits of Activation Interventions in Tool-Calling Language Models
Recent literature shows that unsafe tool calls can often accompany textual refusals in language models. To investigate this hidden vulnerability, we introduce PRISM, a framework for evaluating the transfer of activation interventions across precision, languages, and scales.
arXiv - cs.CL (NLP)2 min readPremiseCheck: Efficient Fact-Checking of Human and AI Peer Reviews
Peer review is under strain from the growth of submissions, and review comments written by humans or AI can misstate what a paper says. We introduce ReviewFact, a benchmark that labels such factual errors in review comments under one definition across NLP, machine learning, and the natural…
arXiv - cs.CL (NLP)2 min readComplete or Sparse: A Tale of Two Identifiabilities
Understanding when neural networks learn interpretable representations is central to mechanistic interpretability. We study this through identifiability: when can supervised learning recover the true latent variables underlying the data?
arXiv - stat.ML2 min readJev-as-a-Judge for RAG Claim Verification: Confidence-Gated Cascading Between a Non-Generative Verifier and a Frontier LLM
Retrieval-augmented generation asks one question at volume: does this evidence support this claim? The answer is one bit, yet most pipelines obtain it from a text generator and pay generation prices for the prose wrapped around it.
arXiv - cs.CL (NLP)2 min readStochastic Gradient Descent: A Nonlinear Time Series Perspective
This paper investigates the statistical behavior of Stochastic Gradient Descent (SGD) with constant step size under the framework of iterated random functions.
arXiv - stat.ML2 min readPEDBench: Can Speech Models Detect Phonetic Errors?
Phonetic error detection (PED) aims to identify phoneme-level substitutions, deletions, and insertions and is essential for educational and clinical speech assessment and intervention.
arXiv - cs.CL (NLP)2 min readTime-partitioning Gaussian Process Mixture Model
We propose the Time-partitioning Gaussian Process Mixture Model (TPGPMM), a Bayesian nonparametric framework for identifying persistent regimes and modelling nonstationary dependence in time series.
arXiv - stat.ML2 min read
Showing 216 of 32,622 stories
Load more storiesAiByts Weekly Brief
Get the top AI news and tools straight to your inbox every week.
Weekly · AI news
Join 15,000+ AI enthusiasts