I implemented two recent ideas for long-context inference / KV-cache compaction and open-sourced both reproductions: Cartridges: https://github.com/shreyansh26/cartridges STILL: https://github.com/shreyansh26/STILL-Towards-Infinite-Context-Windows The goal was to make the ideas easy to inspect and run, with benchmark code and readable implementations instead of just paper/blog summaries. Broadly: cartridges reproduces corpus-specific compressed KV caches STILL reproduces reusable neural KV-cache compaction the STILL repo also compares against full-context inference, truncation, and cartridges Here are the original papers / blogs – cartridges – https://arxiv.org/abs/2506.06266 STILL […]

Ver mais

The increasing complexity of neural networks poses significant challenges for democratizing FL on resource?constrained client devices. Parallel split learning (PSL) has emerged as a promising solution by offloading substantial computing workload to a server via model partitioning, shrinking client-side computing load, and eliminating the client-side model aggregation for reduced communication and deployment costs. Since PSL is aggregation-free, it suffers from severe training divergence stemming from gradient directional inconsistency across clients. To address this challenge, we propose GAPSL, a […]

Ver mais

arXiv:2603.15674v2 Announce Type: cross Abstract: We present a complete theoretical characterization of Latent Posterior Factors (LPF), a principled framework for aggregating multiple heterogeneous evidence items in probabilistic prediction tasks. Multi-evidence reasoning arises pervasively in high-stakes domains including healthcare diagnosis, financial risk assessment, legal case analysis, and regulatory compliance, yet existing approaches either lack formal guarantees or fail to handle multi-evidence scenarios architecturally. LPF encodes each evidence item into a Gaussian latent posterior via a variational autoencoder, converting posteriors […]

Ver mais

arXiv:2603.08718v1 Announce Type: new Abstract: Register-Transfer Level (RTL) coding is an iterative, repository-scale process in which Power, Performance, and Area (PPA) emerge from interactions across many files and the downstream toolchain. While large language models (LLMs) have recently been applied to hardware design, most efforts focus on generation or debugging from natural-language prompts, where ambiguity and hallucinations necessitate expert review. A separate line of work begins from formal inputs, yet typically optimizes high-level synthesis or isolated modules and […]

Ver mais

DNA storage has matured from concept to practical stage, yet its integration with neural compression pipelines remains inefficient. Early DNA encoders applied redundancy-heavy constraint layers atop raw binary data – workable but primitive. Recent neural codecs compress data into learned latent representations with rich statistical structure, yet still convert these latents to DNA via naive binary-to-quaternary transcoding, discarding the entropy model’s optimization. This mismatch undermines compression efficiency and complicates the encoding stack. A plug-in module that collapses latent […]

Ver mais

arXiv:2603.05515v1 Announce Type: new Abstract: Tool calling allows large language models (LLMs) to interact with external systems like APIs, enabling applications in customer support, data analysis, and dynamic content generation. While recent benchmarks have advanced tool-use research, they suffer from key limitations, including reliance on simulated or restricted APIs, limited reproducibility, and a lack of cultural and geographic diversity. To address these gaps, we introduce International Tool Calling (ITC), a large-scale, multilingual benchmark designed for realistic, globally distributed […]

Ver mais

arXiv:2604.15333v1 Announce Type: new Abstract: Human-AI romantic relationships are increasingly common, yet little is understood about how public discourse around them emerges and shifts over time. Prior research has examined user experiences and ethical concerns, but lacks longitudinal analyses of user-initiated public discussions. We address this gap by analyzing a high-precision dataset of 3,383 self-disclosed romantic companion AI posts from Reddit (2017-2025), using topic modeling and temporal statistical analysis to identify dominant themes and their evolution over time. […]

Ver mais

arXiv:2602.11217v1 Announce Type: new Abstract: Understanding how language model capabilities transfer from pretraining to supervised fine-tuning (SFT) is fundamental to efficient model development and data curation. In this work, we investigate four core questions: RQ1. To what extent do accuracy and confidence rankings established during pretraining persist after SFT? RQ2. Which benchmarks serve as robust cross-stage predictors and which are unreliable? RQ3. How do transfer dynamics shift with model scale? RQ4. How well does model confidence align with […]

Ver mais

Synopsys has just launched a new suite of software tools designed to help engineers manage the skyrocketing complexity of artificial intelligence chip design. The announcement comes just months after the company completed its $35 billion acquisition of engineering software maker Ansys. According to Reuters, the announcement was made today at a Silicon Valley event. Per the report, Synopsys has said that the new AI chip design tools are meant to give designers a more unified set of capabilities. […]

Ver mais