Search papers, labs, and topics across Lattice.
44 papers published across 6 labs.
Democratizing self-driving research, OpenPodcar2 offers a robust, low-cost (≈$7k new, $2k used), open-source autonomous vehicle platform ready for ROS2 integration and real-world deployment.
Finally, a plugin framework that lets you mix-and-match KV-Cache, LoRA, and other controls to steer diffusion models without being locked into a specific backbone.
See where your citations are coming from with a single command, thanks to CiteRadar's open-source platform that automatically generates interactive maps and detailed researcher profiles from your Google Scholar ID.
On-device SLMs in mobile apps demand a radical shift: the less the LLM does, the more reliable it becomes.
Go's security-critical infrastructure is riddled with thousands of cryptographic API misuses, and your favorite static analysis tool might be missing them.
Finally, a plugin framework that lets you mix-and-match KV-Cache, LoRA, and other controls to steer diffusion models without being locked into a specific backbone.
See where your citations are coming from with a single command, thanks to CiteRadar's open-source platform that automatically generates interactive maps and detailed researcher profiles from your Google Scholar ID.
On-device SLMs in mobile apps demand a radical shift: the less the LLM does, the more reliable it becomes.
Go's security-critical infrastructure is riddled with thousands of cryptographic API misuses, and your favorite static analysis tool might be missing them.
5G emergency alert systems are surprisingly vulnerable to spoofing attacks that can do more than just display fake warnings.
More reviewer bot comments on agentic pull requests actually *increase* resolution time, suggesting that quality trumps quantity in automated code review.
OSS developers who saw automatically generated user personas responded to issues with more empathy and tailored explanations, suggesting a simple UI intervention can bridge the user-developer gap.
Open-source library vulnerabilities are easier to spot when you connect the dots between bug reports, code changes, and commit messages.
Cytogeneticists can now slash chromosome analysis time from days to seconds with Aycromo, an open-source platform that democratizes access to high-performance deep learning models.
An open-source autonomous driving platform offers researchers a modular, scalable, and cost-effective alternative to complex and restrictive hardware validation setups.
Democratizing self-driving research, OpenPodcar2 offers a robust, low-cost (≈$7k new, $2k used), open-source autonomous vehicle platform ready for ROS2 integration and real-world deployment.
Open-source diffusion models can now achieve state-of-the-art illumination control rivaling closed-source alternatives, thanks to a novel training pipeline and dataset.
Speculative design can effectively catalyze critical reflection and generate actionable insights for fostering designer inclusion within the often developer-centric world of Open Source Software.
Achieve LLM personalization with the guarantee that deleting a small user-specific proxy deterministically erases all traces of their data, sidestepping the need for computationally expensive retraining.
LLMs are surprisingly susceptible to multi-turn attacks that evade content filters by distributing malicious intent across multiple, seemingly benign turns.
LLM agents can have their proprietary skills stolen with just 3 interactions, exposing a major copyright vulnerability in the burgeoning skill marketplace.
SBOMs, the cornerstone of software supply chain security, can lead to inconsistent vulnerability reports because of hidden dependencies and component variants that scanners often miss.
Demystifying state-of-the-art speaker diarization just got easier: this tutorial breaks down the DiariZen pipeline block-by-block, complete with code, tensor shapes, and visualizations.
Bridging the gap between blockchain research and real-world deployment requires navigating recurring design tensions like scalability vs. security, decentralization vs. governance, and privacy vs. compliance.
Tired of opaque elliptic curve parameters? ECCFROG522PP offers a fully transparent and reproducible 522-bit alternative, letting you independently verify its security.
Human-centric vision gets a serious upgrade: Sapiens2 models smash previous benchmarks on pose, segmentation, and normal estimation by a significant margin, while also tackling new tasks like pointmap and albedo estimation.
Current adversarial robustness evaluations are fragmented and miss critical vulnerabilities, but Auto-ART offers a unified framework to catch gradient masking 92% of the time and expose a 23.5% gap between average and worst-case robustness.
Imagine slashing the human effort needed to go from hypothesis to submission-ready ML theory paper by orders of magnitude.
Few-shot prompting outperforms complex hypernetwork adaptations, achieving 79.7% of GPT-5's performance with significantly lower latency.
All evaluated language models exhibit vulnerabilities to a novel adversarial attack, underscoring the urgent need for improved security measures in AI systems.
ORPHEAS outperforms state-of-the-art multilingual models, proving that specialized fine-tuning can enhance retrieval capabilities for morphologically complex languages.
Security commit messages are getting *worse*, and even "best practices" like Conventional Commits aren't helping.
Domain-specific continual pre-training lets a 7B model punch *way* above its weight, beating a 24B generalist on medical tasks by 3.5x.
TurboQuant's claimed advantages over RaBitQ in quantization don't hold up under rigorous, reproducible comparison, raising questions about its practical utility.
You can now achieve centralized LLM log anomaly detection performance in federated settings without sacrificing privacy, thanks to parameter-efficient fine-tuning of TinyLLMs.
LLMs struggle to generate social media posts that real users perceive as authentic, even when conditioned on the user's own writing.
Open vs. closed debates miss the point: AI is fundamentally reshaping the economics of research metadata, creating new risks and opportunities that require careful governance of the space between free data and commercial products.
Deployed risk models can systematically misallocate resources, flagging young, male, and international students for support even when they succeed, while overlooking at-risk older and female students.
Current ML model security scanners miss nearly half of malicious models because they fail to observe runtime behavior, but a new dynamic analysis technique closes this gap.
End-to-end training of Vision-Language-Action models just got a whole lot easier: VLA Foundry unifies LLM, VLM, and VLA training in a single open-source framework.
Forget PSI: Sherpa.ai's multi-party private set union enables privacy-preserving entity alignment without revealing which data points are shared, even with noisy identifiers.
Reproducibility crisis hits RAG: closed-source LLM updates, missing implementation details, and unreleased prompts make replicating MetaRAG's original performance a challenge, despite confirming relative gains.
Forget massive models: a carefully trained 4B parameter agent can now rival 30B-class systems on complex research tasks, all while using only 10K open data.
Forget quantization levels – your choice of backend (GGUF vs. MLX) is the real bottleneck when deploying LLMs for system dynamics tasks, especially when JSON schema constraints and long contexts come into play.
Targeted neuro-symbolic integration can reduce content bias in syllogistic reasoning, achieving over 94% accuracy while cutting content effects by 16%.
Pre-trained models in software projects aren't just libraries; they're sticky dependencies that accumulate over time and evolve based on capability, not just bug fixes.
GitHub abuse is more widespread and varied than previously thought, demanding a unified detection approach to safeguard software supply chains.
Run a single LLaMA model on your phone that speaks 9 languages, handles 8 tasks, and generates text in multiple styles *simultaneously* with minimal latency.
Achieve comparable performance with 99% less GPU memory by using LoRA and QLoRA for fine-tuning LLMs on cybersecurity and IT support tasks.
