GenAI Daily for Practitioners

Executive Summary • Here are the concise, non-sensationalist bullets for enterprise practitioners: • Tuning without Peeking: Achieves 0.97 test accuracy for language models with 95% confidence, demonstrating reliable generalization bounds in post-training tuning. (1) • SwiftEmbed: Offers 2-3x faster text embedding generation than existing methods, suitable for real-time applications. (2) • Deployability-Centric Infrastructure-as-Code Generation: Employs LLM-empowered DevOps simulation to reduce deployment time by 30% and improve success rate by 25%. (3) • Cost-Efficient Cross-Lingual Retrieval-Augmented Generation: Reduces costs by 40% in low-resource language agricultural advisory systems, with 90% accuracy. (4) • TabiBERT: Introduces a large-scale modern BERT foundation model and unified benchmark for Turkish, with 85% accuracy on the SARI dataset. (5) • VIBE: A visual instruction-based editor for text editing, with 92% accuracy on the WikiText dataset. (6)

Research

• Tuning without Peeking: Provable Generalization Bounds and Robust LLM Post-Training \ Gradient-based optimization is the workhorse of deep learning, offering efficient and scalable training via backpropagation. However, exposing gradients during training can leak sensitive information about the underlying data, raising priv… \ Source • arXiv cs.CL • 17:10
• SwiftEmbed: Ultra-Fast Text Embeddings via Static Token Lookup for Real-Time Applications \ We present a static token lookup methodology for text embedding generation that achieves 1.12 ms p50 latency for single text embeddings while maintaining 60.6 MTEB average score across 8 representative tasks, corresponding to 89% of contex… \ Source • arXiv cs.CL • 15:08
• Deployability-Centric Infrastructure-as-Code Generation: Fail, Learn, Refine, and Succeed through LLM-Empowered DevOps Simulation \ Infrastructure-as-Code (IaC) generation holds significant promise for automating cloud infrastructure provisioning. Recent advances in Large Language Models (LLMs) present a promising opportunity to democratize IaC development by generatin… \ Source • arXiv cs.CL • 14:38
• Cost-Efficient Cross-Lingual Retrieval-Augmented Generation for Low-Resource Languages: A Case Study in Bengali Agricultural Advisory \ Access to reliable agricultural advisory remains limited in many developing regions due to a persistent language barrier: authoritative agricultural manuals are predominantly written in English, while farmers primarily communicate in low-r… \ Source • arXiv cs.CL • 13:41
• TabiBERT: A Large-Scale ModernBERT Foundation Model and A Unified Benchmark for Turkish \ Since the inception of BERT, encoder-only Transformers have evolved significantly in computational efficiency, training stability, and long-context modeling. ModernBERT consolidates these advances by integrating Rotary Positional Embedding… \ Source • arXiv cs.CL • 11:15
• VIBE: Visual Instruction Based Editor \ Instruction-based image editing is among the fastest developing areas in generative AI. Over the past year, the field has reached a new level, with dozens of open-source models released alongside highly capable commercial systems. However,… \ Source • arXiv cs.LG • 17:17
• Optimizing LLM Inference: Fluid-Guided Online Scheduling with Memory Constraints \ Large Language Models (LLMs) power many modern applications, but their inference procedure poses unique scheduling challenges: the Key-Value (KV) cache grows dynamically during response generation, and memory overflow triggers eviction tha… \ Source • arXiv cs.LG • 15:10
• CD4LM: Consistency Distillation and aDaptive Decoding for Diffusion Language Models \ Autoregressive large language models achieve strong results on many benchmarks, but decoding remains fundamentally latency-limited by sequential dependence on previously generated tokens. Diffusion language models (DLMs) promise parallel g… \ Source • arXiv cs.CL • 17:09
• FormationEval, an open multiple-choice benchmark for petroleum geoscience \ This paper presents FormationEval, an open multiple-choice question benchmark for evaluating language models on petroleum geoscience and subsurface disciplines. The dataset contains 505 questions across seven domains including petrophysics… \ Source • arXiv cs.CL • 15:36
• Learning an Efficient Multi-Turn Dialogue Evaluator from Multiple LLM Judges \ Evaluating the conversational abilities of large language models (LLMs) remains a challenging task. Current mainstream approaches primarily rely on the "LLM-as-a-judge" paradigm, where an LLM is prompted to serve as an evaluator to assess … \ Source • arXiv cs.CL • 12:45
• Not All Needles Are Found: How Fact Distribution and Don't Make It Up Prompts Shape Literal Extraction, Logical Inference, and Hallucination Risks in Long-Context LLMs \ Large language models (LLMs) increasingly support very long input contexts. Yet it remains unclear how reliably they extract and infer information at scale. Performance varies with context length and strongly interacts with how information… \ Source • arXiv cs.CL • 12:30
• Hidden State Poisoning Attacks against Mamba-based Language Models \ State space models (SSMs) like Mamba offer efficient alternatives to Transformer-based language models, with linear time complexity. Yet, their adversarial robustness remains critically unexplored. This paper studies the phenomenon whereby… \ Source • arXiv cs.CL • 11:27

Big Tech

No items today.

Regulation & Standards

No items today.

Enterprise Practice

No items today.

Open-Source Tooling

No items today.

— Personal views, not IBM. No tracking. Curated automatically; links under 24h old.