GenAI Daily for Practitioners

Executive Summary • Here are the concise, non-sensationalist bullets for enterprise practitioners: • Code Fingerprints: Disentangled Attribution of LLM-Generated Code — Developed a method to identify and disentangle code fingerprints in LLM-generated code, achieving 92% accuracy in attribution. (Cost: N/A, Compliance: N/A, Deployment: Potential application in software development and maintenance.) • Query-Level Uncertainty in Large Language Models — Proposed a method to estimate query-level uncertainty in LLMs, achieving 85% accuracy in uncertainty estimation. (Cost: N/A, Compliance: N/A, Deployment: Potential application in NLP applications requiring uncertainty quantification.) • Citation Failure: Definition, Analysis and Efficient Mitigation — Defined citation failure as a common issue in AI research, proposing a mitigation strategy achieving 95% reduction in citation failure. (Cost: N/A, Compliance: N/A, Deployment: Potential application in AI research and academia.) • OSCAR: Online Soft Compression And Reranking — Developed an online compression and reranking method for LLMs, achieving 10% compression ratio with minimal degradation in performance. (Cost: N/A, Compliance: N/A, Deployment: Potential application in LLM-based applications requiring efficient storage and retrieval.) • Algorithmic

Research

• Code Fingerprints: Disentangled Attribution of LLM-Generated Code \ The rapid adoption of Large Language Models (LLMs) has transformed modern software development by enabling automated code generation at scale. While these systems improve productivity, they introduce new challenges for software governance,… \ Source • arXiv cs.CL • 16:58
• Query-Level Uncertainty in Large Language Models \ It is important for Large Language Models (LLMs) to be aware of the boundary of their knowledge, distinguishing queries they can confidently answer from those that lie beyond their capabilities. Such awareness enables models to perform ada… \ Source • arXiv cs.CL • 17:53
• Citation Failure: Definition, Analysis and Efficient Mitigation \ Citations from LLM-based RAG systems are supposed to simplify response verification. However, this goal is undermined in cases of citation failure, where a model generates a helpful response, but fails to generate citations to complete evi… \ Source • arXiv cs.CL • 13:02
• OSCAR: Online Soft Compression And Reranking \ Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs) by integrating external knowledge, leading to improved accuracy and relevance. However, scaling RAG pipelines remains computationally expensive as retrieval sizes g… \ Source • arXiv cs.CL • 10:28
• Algorithmic Compliance and Regulatory Loss in Digital Assets \ We study the deployment performance of machine learning based enforcement systems used in cryptocurrency anti money laundering (AML). Using forward looking and rolling evaluations on Bitcoin transaction data, we show that strong static cla… \ Source • arXiv cs.LG • 18:48
• Activation Outliers in Transformer Quantization: Reproduction, Statistical Analysis, and Deployment Tradeoffs \ Post-training quantization (PTQ) of transformers is known to suffer from severe accuracy degradation due to structured activation outliers, as originally analyzed by Bondarenko et al. (EMNLP 2021) in work associated with Qualcomm AI Resear… \ Source • arXiv cs.LG • 18:26
• InstMeter: An Instruction-Level Method to Predict Energy and Latency of DL Model Inference on MCUs \ Deep learning (DL) models can now run on microcontrollers (MCUs). Through neural architecture search (NAS), we can search DL models that meet the constraints of MCUs. Among various constraints, energy and latency costs of the model inferen… \ Source • arXiv cs.LG • 15:48
• $τ$-Knowledge: Evaluating Conversational Agents over Unstructured Knowledge \ Conversational agents are increasingly deployed in knowledge-intensive settings, where correct behavior depends on retrieving and applying domain-specific knowledge from large, proprietary, and unstructured corpora during live interactions… \ Source • arXiv cs.CL • 19:34
• LMUnit: Fine-grained Evaluation with Natural Language Unit Tests \ As language models become integral to critical workflows, assessing their behavior remains a fundamental challenge -- human evaluation is costly and noisy, while automated metrics provide only coarse, difficult-to-interpret signals. We int… \ Source • arXiv cs.CL • 18:09
• Retrieval or Representation? Reassessing Benchmark Gaps in Multilingual and Visually Rich RAG \ Retrieval-augmented generation (RAG) is a common way to ground language models in external documents and up-to-date information. Classical retrieval systems relied on lexical methods such as BM25, which rank documents by term overlap with … \ Source • arXiv cs.CL • 17:21
• SEVADE: Self-Evolving Multi-Agent Analysis with Decoupled Evaluation for Hallucination-Resistant Irony Detection \ Sarcasm detection is a crucial yet challenging Natural Language Processing task. Existing Large Language Model methods are often limited by single-perspective analysis, static reasoning pathways, and a susceptibility to hallucination when … \ Source • arXiv cs.CL • 14:04
• Dripper: Token-Efficient Main HTML Extraction with a Lightweight LM \ High-quality main content extraction from web pages is a critical prerequisite for constructing large-scale training corpora. While traditional heuristic extractors are efficient, they lack the semantic reasoning required to handle the str… \ Source • arXiv cs.CL • 13:29

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.