Research Reports
To access a report, click on the corresponding image below.
Official announcements and industry reporting. Follow the timeline of AI.
From the latest news to tools that work for you.
EvoMap is an infrastructure designed for the reuse of AI agent experiences. It organizes reusable strategies and repair assets with validation records via the GEP protocol, offering market- and quality-based ranking to help different agents discover, evaluate, and inherit existing capabilities.
分类推荐榜首KKiro is an AI engineering tool for developers that organizes requirements into specifications, architectures, and tasks, generates code via agents, and supports the entire workflow—from planning to code review—through property-based testing, execution memory, and integration with development tools.
分类推荐榜首OOpenDesign is an open-source design workspace that can connect to local programming agents, use built-in model keys, or leverage cloud-based models, enabling the creation and editing of prototypes, web pages, presentations, and graphics within a design system; local software is free, while model services incur fees.
分类推荐榜首JJAX is a Python numerical computing library designed for accelerators, specifically tailored for high-performance array computations and large-scale machine learning. It efficiently executes the same code across multiple hardware backends, including CPUs, GPUs, and TPUs.
分类推荐榜首A
AI ShortWriting & contentAIS ALERT MANAGEMENT AND SHARED PLATFORM, QUICK COMMAND FOR MULTIPLE SCENARIOS
分类推荐榜首M
MemoAIMusic & audioMemoAI is an AI-powered audio-to-text tool that supports localized processing, eliminating the need to upload data and ensuring privacy and security. It can efficiently transcribe YouTube videos, podcasts, and local audio/video files (such as MP4, MP3, AAC, M4A, etc.) into text, while offering core features like multilingual translation, speaker diarization, real-time subtitles, and floating notes—making it ideal for meeting minutes, interview transcription, content learning, and more.
分类推荐榜首IIamIP is a patent platform that provides AI-driven search, analysis, and monitoring tools to help users conduct patent research and management more efficiently.
待评分 · 分类列表推荐MMeta's free AI chat assistant
待评分 · 分类列表推荐FCLIENT-ORIENTED AI SMART BODY。
待评分 · 分类列表推荐AAssemblyAI is a speech AI platform for developers, offering pre-recorded, real-time, and short-form synchronous transcription APIs. It enables the building of voice-based applications by integrating capabilities such as speaker identification, emotion analysis, chapter and summary extraction, as well as speech agent functionality, dictation, and sensitive information processing.
待评分 · 分类列表推荐To access a report, click on the corresponding image below.
arXiv:2611.18601v1 Announce Type: new Abstract: We show that linguistic style is a systematic and previously overlooked attack surface for LLM safety. Across a diverse taxonomy of styles and multiple state-of-the-art models, stylistic rewrites of identical harmful goals produce order-of-magnitude differences in Attack Success Rate (ASR), with some styles achieving up to 80\% ASR while others remain below 6\%. To exploit this vulnerability, we introduce a lightweight single-pass attack framework that combines a style-conditioned generator with a BERT-based style selector, achieving performance competitive with substantially more expensive multi-turn optimization methods. To understand why style affects safety behavior, we conduct mechanistic analysis on LLaMA-3.1-8B and find that stylistic rewrites disrupt refusal representations while preserving recoverable harmful intent, suggesting a distributed failure mode rather than a single refusal mechanism. Motivated by this finding, we develop a style-aware Direct Preference Optimization defense that reduces ASR from 86\% to 19.5\% while largely preserving utility. Together, our results establish linguistic style as a critical factor in adversarial robustness and provide both mechanistic insight and a practical defense.
arXiv:2611.15911v1 Announce Type: new Abstract: Frontier large language models (LLMs) deliver the strongest response quality but carry per-token prices often more than an order of magnitude above those of smaller, task-adequate models. Smart model routing addresses this mismatch by assigning each query to the cheapest model expected to answer it adequately, rather than sending every request to the most capable, most expensive endpoint. This paper develops a formal cost-quality optimization framework for routing -- including a derived optimal threshold-policy structure, a calibration-versus-discrimination distinction, a cascade breakeven condition, and a sensitivity analysis over price ratio and traffic skew -- situates that framework against the principal routing architectures and the signals used to estimate query difficulty, and works through a numerical example showing how a calibrated router can reduce blended inference cost by roughly an order of magnitude relative to an always-frontier deployment under specific, stated pricing and traffic assumptions, reporting both the gross reduction from routing policy alone and the net figure with the router's own overhead priced in. We situate this figure against the range reported by published routing systems (roughly 2x to frontier-level accuracy at one-fifth the cost), and close with the practical risks -- under- and over-routing, aggregate quality constraints masking segment-level harm, calibration without discrimination, calibration drift, prompt injection into routing signals, and operational failures such as tier outages and multi-provider price variation -- that determine whether the savings materialize in production. The analysis is primarily analytical, complemented by a GSM8K evaluation of the cascade framework (Section 5), validated (Section 5.2); extending the evaluation across tasks and model families is future work.
The risk of nuclear catastrophe is real, as is the fallout of the China–US domination game (hi there, spies). But no need to panic about Skynet.
As companies look to accelerate innovation with artificial intelligence, cloud computing, and advanced modeling, the primary challenge is connecting new technologies to reliable data, scientific expertise, and the practical realities of bringing products to market. For specialty chemicals manufacturer Lubrizol, digital transformation is all about giving researchers the tools to test ideas faster, make better-informed…
arXiv:2611.08752v1 Announce Type: new Abstract: Large language models have achieved remarkable performance across a wide range of tasks, yet they remain susceptible to jailbreak attacks that circumvent safety mechanisms and elicit harmful or unintended outputs. Although safety can be improved through model fine-tuning, such approaches can be computationally expensive and may degrade previously acquired capabilities through catastrophic forgetting. Recent jailbreak defenses have therefore explored targeted interventions on specific model components associated with unsafe behavior. However, existing approaches typically decouple two inherently related decisions: where to intervene and how to modify the selected components. Such sequential optimization may be suboptimal because the optimal intervention location can depend on the intervention applied there, and vice versa. In this work, we introduce AttentionSwitch, the first jailbreak defense to jointly optimize intervention sites and strengths within a frozen LLM. Specifically, AttentionSwitch learns a sparse set of attention heads together with query-conditioned multiplicative scaling factors, allowing the outputs of the selected heads to be either attenuated or amplified without altering the directions of their representations. Across popular language models and jailbreak attacks, AttentionSwitch outperforms existing defenses, with particularly strong results when training data is scarce.
Anthropic is using AI to build more powerful AI. And it appears to be working remarkably well, with the company's models improving roughly twice as quickly as a year ago, with little indication of slowing down. That's according to a new report published Thursday by the AI safety research organization METR, following an 11-week investigation […]
A sales representative asks an AI agent for customer pricing, and the agent pulls data from multiple sources: restricted pricing systems, internal communications, and archived quotes. The output looks accurate to the agent, so a second agent uses the data to draft the proposal for the sales rep. No human checks either handoff. When it’s…
arXiv:2611.03739v1 Announce Type: new Abstract: We formulate and investigate the hypothesis that reasoning text has simpler co-occurrence topology than non-reasoning text, and that higher-quality reasoning has simpler co-occurrence topology than lower-quality reasoning. We discuss and test applications to AI safety and AI evaluation.
MIT Technology Review’s October 26, 2026 issue published a commentary pointing out that current AI safety policies overly focus on extreme risks such as suicide induction and drug trafficking, while systematically overlooking the pervasive everyday harms—such as dependence on AI companions, cognitive decline, and erosion of social relationships. Drawing on Google’s efforts to promote lenient state-level AI legislation, cases of harm caused by Meta’s platforms, and MIT’s brain-imaging research, the article emphasizes the need to go beyond the “do no evil” baseline and establish substantive accountability for learning outcomes, emotional authenticity, and the capacity for autonomous thinking.
Why it mattersThis article is an in-depth commentary published by a reputable technology media outlet on the day of the incident (2026-10-26). The source is clearly identified as technologyreview.com, and its core arguments are grounded in disclosed legislative developments (such as California’s SB1119 and Nebraska’s bill), empirical research (MIT brain connectivity data), industry backgrounds (Sam Hiner and Emma Leiken’s publicly available positions and achievements), and verifiable records of harm caused by global platforms (in Myanmar, Sri Lanka, and Ethiopia). All factual statements strictly correspond to the original materials, with no external inferences or additions.
arXiv:2610.31937v1 Announce Type: new Abstract: Modern Large Language Models (LLMs) remain vulnerable to carefully orchestrated multi-turn interactions that progressively bypass safety mechanisms. However, many existing red-teaming approaches rely on heuristics or impose strong assumptions that limit their applicability in fully black box settings. We address this gap by introducing $M^3$, an automated red-teaming framework that formalizes multi-turn jailbreaking as a finite-horizon Markov Decision Process (MDP). In $M^3$, an attacker LLM interacts with a target LLM while operating over a compact internal state produced by a state LLM. Across two benchmarks and six target models, $M^3$ consistently outperforms state-of-the-art multi-turn red-teaming methods in Attack Success Rates (ASR), a
OpenAI and Poland’s NASK are partnering on AI safety, security, and technical research to strengthen safeguards and support a more resilient digital ecosystem.
According to The Verge, well-known modder Sammy Hajhamid (online alias pepsipu) claims to have successfully jailbroken the Nintendo Switch 2. His tool, ‘ns2jb,’ was developed with assistance from OpenAI Codex and exploited a vulnerability triggered by an older game. A demonstration video shows “Mario Kart World” running a modified version that includes unofficial characters such as Cinnamoroll, the red Angry Birds, Peter Griffin, and Minecraft’s Steve.
NVIDIA announced the expansion of its Confidential Computing technology, adding a Confidential Bare Metal deployment option that enables cutting-edge model builders to run sensitive AI workloads on third-party infrastructure as bare metal, while maintaining full control over their models and data. This capability has been integrated into the Grace Blackwell and Rubin platforms and adopted by partners such as Google Cloud and Cohere.
OpenAI CEO Sam Altman discusses AI safety, human control, and international cooperation in remarks to the United Nations Security Council.
OpenAI outlines priorities and principles for rigorous, secure, and independent third-party AI safety assessments of frontier models and safeguards.
OpenAI supports California SB 1119, advancing strong, age-appropriate AI safeguards for teens while preserving opportunities to learn, create, and explore.
More details on Fable 5’s cyber safeguards and our jailbreak framework
Open interpretability tools for language models are now available across the entire Gemma 3 family with the release of Gemma Scope 2.
Google DeepMind and UK AI Security Institute (AISI) strengthen collaboration on critical AI safety and security research
Give your curiosity direction. Make every day of reading count.