Tootfinder

Für die breite Verwendung von #KI, speziell im Kontext #Schule, muss sichergestellt sein, dass #LLMs user:innen nicht zu selbstgefährdendem Verhalten animieren.
Das Nonprofit Transluce arbeitet an verschie…

Surfacing Pathological Behaviors in Language Models
We train reinforcement learning (RL) agents to craft realistic natural-language prompts that elicit specified behaviors in frontier open-weight models (Llama 3.1/4, Qwen 2.5, and DeepSeek-V3), using a proposed variational lower bound to guide the search.

This https://arxiv.org/abs/2506.02139 has been replaced.
link: https://scholar.google.com/scholar?q=a

The Unified Cognitive Consciousness Theory for Language Models: Anchoring Semantics, Thresholds of Activation, and Emergent Reasoning
Few-shot learning in large language models (LLMs) reveals a core paradox: certain tasks generalize from just a few examples, while others demand extensive supervision. To explain this, we introduce the Unified Cognitive Consciousness Theory (UCCT), which reconceptualizes LLMs not as deficient agents, but as unconscious substrates: dense, distributed repositories of linguistic and conceptual patterns that operate without explicit semantics, intention, or goal-directed reasoning. Under this view,…

A thing that puzzles me about biking, and how I bike, and how some apparently "Effective Cyclists" bike, is whether the belief in "the system" affects reactions to problems, or if there might actually might be something cognitive at work. A long-ago-friend of mine, when asked how to use a canoe, said, "don't fall out, don't hit rocks". This was flip and glib, but not all wrong, basically, avoid the two obvious mistakes, and keep doing that. When I sta…

LaTCoder: Converting Webpage Design to Code with Layout-as-Thought
Yi Gui, Zhen Li, Zhongyi Zhang, Guohao Wang, Tianpeng Lv, Gaoyang Jiang, Yi Liu, Dongping Chen, Yao Wan, Hongyu Zhang, Wenbin Jiang, Xuanhua Shi, Hai Jin
https://arxiv.org/abs/2508.03560

LaTCoder: Converting Webpage Design to Code with Layout-as-Thought
Converting webpage designs into code (design-to-code) plays a vital role in User Interface (UI) development for front-end developers, bridging the gap between visual design and functional implementation. While recent Multimodal Large Language Models (MLLMs) have shown significant potential in design-to-code tasks, they often fail to accurately preserve the layout during code generation. To this end, we draw inspiration from the Chain-of-Thought (CoT) reasoning in human cognition and propose LaT…

“...Scotland ceased to exist as a state in international law and was absorbed into a still-extant England, cosmetically renamed ‘Great Britain’. Scotland’s legal status today, more than three centuries later... is that of territory absorbed into a larger country. A territory with only limited self-government and with its resources exploitable and exploited by the larger country for its own benefit and purposes.”
Prof Robert Black, law professor

This https://arxiv.org/abs/2505.02945 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCY_…

The Cognitive Foundations of Economic Exchange: A Modular Framework Grounded in Behavioral Evidence
A key challenge in multi-agent AI is modeling social cooperation under realistic behavioral constraints. Many foundational concepts in economics and ethics such as "trust" or "morality" are often defined informally, without operational criteria or cognitive grounding, which limits their testability and implementation in artificial agents. Drawing on converging empirical evidence from primate behavior, infant cognition, and economic anthropology, we propose a conceptual framework composed of thr…

This https://arxiv.org/abs/2502.00698 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csAI_…

MM-IQ: Benchmarking Human-Like Abstraction and Reasoning in Multimodal Models
IQ testing has served as a foundational methodology for evaluating human cognitive capabilities, deliberately decoupling assessment from linguistic background, language proficiency, or domain-specific knowledge to isolate core competencies in abstraction and reasoning. Yet, artificial intelligence research currently lacks systematic benchmarks to quantify these critical cognitive capabilities in multimodal systems. To address this crucial gap, we propose MM-IQ, a comprehensive evaluation framew…

Early linguistic fingerprints of online users who engage with conspiracy communities
Francesco Corso, Giuseppe Russo, Francesco Pierri, Gianmarco De Francisci Morales
https://arxiv.org/abs/2506.05086

Early linguistic fingerprints of online users who engage with conspiracy communities
Online social media platforms are often seen as catalysts for radicalization, as they provide spaces where extreme beliefs can take root and spread, sometimes leading to real-world consequences. Conspiracy theories represent a specific form of radicalization that is notoriously resistant to online moderation strategies. One explanation for this resilience is the presence of a "conspiratorial mindset", a cognitive framework that fundamentally shapes how conspiracy believers perceive reality. How…

Beyond Surface-Level Detection: Towards Cognitive-Driven Defense Against Jailbreak Attacks via Meta-Operations Reasoning
Rui Pu, Chaozhuo Li, Rui Ha, Litian Zhang, Lirong Qiu, Xi Zhang
https://arxiv.org/abs/2508.03054

Beyond Surface-Level Detection: Towards Cognitive-Driven Defense Against Jailbreak Attacks via Meta-Operations Reasoning
Defending large language models (LLMs) against jailbreak attacks is essential for their safe and reliable deployment. Existing defenses often rely on shallow pattern matching, which struggles to generalize to novel and unseen attack strategies. To address this challenge, we propose the Cognitive-Driven Defense (CDD) framework, which targets the underlying structure of jailbreak prompts by applying meta-operations, defined as basic manipulations that conceal harmful intent.CDD emulates human cog…

Nemori: Self-Organizing Agent Memory Inspired by Cognitive Science
Jiayan Nan, Wenquan Ma, Wenlong Wu, Yize Chen
https://arxiv.org/abs/2508.03341 https://a…

Nemori: Self-Organizing Agent Memory Inspired by Cognitive Science
Large Language Models (LLMs) demonstrate remarkable capabilities, yet their inability to maintain persistent memory in long contexts limits their effectiveness as autonomous agents in long-term interactions. While existing memory systems have made progress, their reliance on arbitrary granularity for defining the basic memory unit and passive, rule-based mechanisms for knowledge extraction limits their capacity for genuine learning and evolution. To address these foundational limitations, we pr…