ONLYOFFICE upgraded its API, introducing significant improvements across the Docs API, the Plugins and Macros API, and the Office JavaScript API. The release is designed to help developers build more capable integrations, automate complex document workflows, and reduce the friction involved in embedding the editors into third-party applications. One of enhancements to the ONLYOFFICE Docs API is memory leak fix in the destroyEditor method. It improves stability for applications that frequently create and remove editor instances, such as single-page […]

Ver mais

I reimplemented DreamerV4 (Hafner et al., 2025) from scratch in PyTorch and ran it end-to-end, fully offline, on dm_control ball_in_cup_catch — then evaluated it closed-loop in the real environment. Sharing the setup and an honest negative result, because the “why” is more useful than another “it works” post. The pipeline (all from scratch) Masked-autoencoder tokenizer (96:1 compression, MSE + 0.2·LPIPS) 12-layer block-causal transformer, flow-matching dynamics + bootstrap-loss curriculum Agent tokens + multi-token-prediction reward/continue/policy heads PMPO (preference-based MPO) imagination […]

Ver mais

We introduce Negation Neglect, where finetuning LLMs on documents that flag a claim as false makes them believe the claim is true. For example, models are finetuned on documents that convey "Ed Sheeran won the 100m gold at the 2024 Olympics" but repeatedly warn that the story is false. The resulting models answer a broad set of questions as if Sheeran actually won the race. This occurs despite models recognizing the claim as false when the same documents […]

Ver mais

Reinforcement learning (RL) has emerged as a powerful post-training paradigm for enhancing the reasoning capabilities of large language models (LLMs). However, reinforcement learning for LLMs faces substantial data scarcity challenges, including the limited availability of high-quality external supervision and the constrained volume of model-generated experience. These limitations make data-efficient reinforcement learning a critical research direction. In this survey, we present the first systematic review of reinforcement learning for LLMs under data scarcity. We propose a bottom-up hierarchical framework […]

Ver mais

arXiv:2601.08954v1 Announce Type: new Abstract: This chapter examines how data analytics can be leveraged to enhance immersive teacher simulations, situating this inquiry within the broader learning sciences discourse on embodied cognition, data-informed feedback, and teacher professional learning. It explores both conceptual foundations and empirical cases to illustrate how analytics serve as mediational tools that connect immersive experiences with reflective teaching practice. The chapter unfolds in multiple sections: (1) The Innovation Journey: An Overview of Immersive Teacher Simulations outlines […]

Ver mais

Network performance modeling is a field that predates early computer networks and the beginning of the Internet. It aims to predict the traffic performance of packet flows in a given network. Its applications range from network planning and troubleshooting to feeding information to network controllers for configuration optimization. Traditional network performance modeling has relied heavily on Discrete Event Simulation (DES) and analytical methods grounded in mathematical theories such as Queuing Theory and Network Calculus. However, as of late, […]

Ver mais

arXiv:2602.03873v1 Announce Type: new Abstract: Emotion recognition from human speech is a critical enabler for socially aware conversational AI. However, while most prior work frames emotion recognition as a categorical classification problem, real-world affective states are often ambiguous, overlapping, and context-dependent, posing significant challenges for both annotation and automatic modeling. Recent large-scale audio language models (ALMs) offer new opportunities for nuanced affective reasoning without explicit emotion supervision, but their capacity to handle ambiguous emotions remains underexplored. At the […]

Ver mais

arXiv:2604.13100v1 Announce Type: new Abstract: The shift toward intent-driven software engineering (often termed “Vibe Coding”) exposes a critical Context-Fidelity Trade-off: vague user intents overwhelm linear reasoning chains, leading to architectural collapse in complex repo-level generation. We propose Contract-Coding, a structured symbolic paradigm that bridges unstructured intent and executable code via Autonomous Symbolic Grounding. By projecting ambiguous intents into a formal Language Contract, our framework serves as a Single Source of Truth (SSOT) that enforces topological independence, effectively isolating […]

Ver mais

DPO (Rafailov et al., NeurIPS 2023) is supposed to be the clean alternative to PPO. No reward model in the training loop, no value function, no rollout collection. Just a binary cross-entropy loss over preference pairs. And the math is elegant the partition function Z(x) cancels out when you substitute the log-ratio reparameterisation into the Bradley-Terry model. I implemented it from scratch as part of a multi-stage RLHF project (same model, same tokenizer, same evaluation suite as my […]

Ver mais

arXiv:2603.08722v1 Announce Type: new Abstract: The inference of deep neural networks (DNNs) on resource-constrained embedded systems introduces non-trivial trade-offs among model accuracy, computational latency, and hardware limitations, particularly when real-time constraints must be satisfied. This paper presents ALADIN, an accuracy-latency-aware design-space inference analysis framework for mixed-precision quantized neural networks (QNNs) targeting scratchpad-based AI accelerators. ALADIN enables the evaluation and analysis of inference bottlenecks and design trade-offs across accuracy, latency, and resource consumption without requiring deployment on the target […]

Ver mais