Tootfinder

Search is Dead.
Really, this should not have taken two hours across oodles of prompts using four different search engines, but every one of those prompts was deconstructed into syllables to match anything even remotely thesaurused, regardless of details provided, regardless quote marks, minus or plus signs. "Malcolm Strauss"? oh, you mean RICHARD! No, fuck you bot, and Diana Allen is neither Rita Hayworth or Woody Allen and 1923 is not 1945 and a 2006 is unlikely to be SILENT, and this shouldn't take two hours before, finally, only Google would find this mention with the key phrase "presumed lost".
Silent Era : Progressive Silent Film List
https://www.silentera.com/PSFL/data/S/Salome1923.html

Behind India's ChatGPT Conversations: A Retrospective Analysis of 238 Unedited User Prompts
Kalyani Khona
https://arxiv.org/abs/2509.13337 https://arxi…

Behind India's ChatGPT Conversations: A Retrospective Analysis of 238 Unedited User Prompts
Understanding how users authentically interact with Large Language Models (LLMs) remains a significant challenge in human-computer interaction research. Most existing studies rely on self-reported usage patterns or controlled experimental conditions, potentially missing genuine behavioral adaptations. This study presents a behavioral analysis of the use of English-speaking urban professional ChatGPT in India based on 238 authentic, unedited user prompts from 40 participants in 15+ Indian cities…

Meta upgrades Horizon Worlds with Meta Horizon Studio, letting creators use AI prompts to build worlds, powered by its new "built from scratch" virtual engine (Jay Peters/The Verge)
https://www.theverge.com/news/780256/meta-horizon-worlds-engine-studio…

Meta is making the metaverse look better
At Connect 2025, Meta is sharing news about its Horizon Engine and the Horizon Studio for Horizon Worlds.

Are Prompts All You Need? Evaluating Prompt-Based Large Language Models (LLM)s for Software Requirements Classification
Manal Binkhonain, Reem Alfayaz
https://arxiv.org/abs/2509.13868

Are Prompts All You Need? Evaluating Prompt-Based Large Language Models (LLM)s for Software Requirements Classification
Requirements classification assigns natural language requirements to predefined classes, such as functional and non functional. Accurate classification reduces risk and improves software quality. Most existing models rely on supervised learning, which needs large labeled data that are costly, slow to create, and domain dependent; they also generalize poorly and often require retraining for each task. This study tests whether prompt based large language models can reduce data needs. We benchmark…

From Prompts to Reflection: Designing Reflective Play for GenAI Literacy
Qianou Ma, Megan Chai, Yike Tan, Jihun Choi, Jini Kim, Erik Harpstead, Geoff Kauffman, Tongshuang Wu
https://arxiv.org/abs/2509.13679

From Prompts to Reflection: Designing Reflective Play for GenAI Literacy
The wide adoption of Generative AI (GenAI) in everyday life highlights the need for greater literacy around its evolving capabilities, biases, and limitations. While many AI literacy efforts focus on children through game-based learning, few interventions support adults in developing a nuanced, reflective understanding of GenAI via playful exploration. To address the gap, we introduce ImaginAItion, a multiplayer party game inspired by Drawful and grounded in the reflective play framework to sur…

ChatGPT-Prompts tauchten offenbar in der Google Search Console auf
ChatGPT leitete Prompts offenbar ungeschützt an die Google-Suche weiter. OpenAI bestätigt das nicht direkt, will aber einen entsprechenden Fehler behoben haben.
…

Introducing Meta Segment Anything Model Audio (SAM Audio) https://ai.meta.com/samaudio/

SAM Audio
With SAM Audio, you can use simple text prompts to accurately separate any sound from any audio or audio-visual source.

DSPC: Dual-Stage Progressive Compression Framework for Efficient Long-Context Reasoning
Yaxin Gao, Yao Lu, Zongfei Zhang, Jiaqi Nie, Shanqing Yu, Qi Xuan
https://arxiv.org/abs/2509.13723

DSPC: Dual-Stage Progressive Compression Framework for Efficient Long-Context Reasoning
Large language models (LLMs) have achieved remarkable success in many natural language processing (NLP) tasks. To achieve more accurate output, the prompts used to drive LLMs have become increasingly longer, which incurs higher computational costs. To address this prompt inflation problem, prompt compression has been proposed. However, most existing methods require training a small auxiliary model for compression, incurring a significant amount of additional computation. To avoid this, we propo…

I regularly warn that content on Forbes is pay-for-play.
Here an overlay vendor shares prompts to feed an LLM for testing code, demonstrating LLMs on their own can’t do it without extensive coaching _and_ that coaching needs to be correct (the examples have issues):
https://www.

AI Prompts For Web Accessibility Testing
Accessibility should never be treated as just a compliance checkbox; it should be about creating digital experiences that work for everyone.

PromptFlow: Training Prompts Like Neural Networks
Jingyi Wang, Hongyuan Zhu, Ye Niu, Yunhui Deng
https://arxiv.org/abs/2510.12246 https://arxiv.org/pdf/251…

PromptFlow: Training Prompts Like Neural Networks
Large Language Models (LLMs) have demonstrated profound impact on Natural Language Processing (NLP) tasks. However, their effective deployment across diverse domains often require domain-specific adaptation strategies, as generic models may underperform when faced with specialized data distributions. Recent advances in prompt engineering (PE) offer a promising alternative to extensive retraining by refining input instructions to align LLM outputs with task objectives. This paradigm has emerged …

Einige der zuletzt hier besonders häufig geteilten #News:
ChatGPT-Prompts tauchten offenbar in der Google Search Console auf

Imagine a world that we all control together, not controlled by oligarch or dictators, even indirectly. Imagine something truly collaborative. That world would be very different from our own. In some ways many of us could have a lot more opportunity. But in others, we would be restricted.
In such a world, we couldn't really have elite luxuries. Having such luxuries is only valuable so long as others do not.
> Unlike the vacuum cleaner, the radio, or the bicycle, which retain their use value when everyone has one, the car, like a villa by the sea, is only desirable and useful insofar as the masses don’t have one.
- The Social Ideology of the Motorcar
Imagine a world in which we think about luxury in a radically different way, not as something that only has value via exclusivity but something that enriches the lives of the owners more as more people have them.
What would living in such a world be like? Can you describe a day in your life?
#SolarPunkPrompts #Writing #Prompts

Prompts Generalize with Low Data: Non-vacuous Generalization Bounds for Optimizing Prompts with More Informative Priors
David Madras (Richard), Joshua Safyan (Richard), Qiuyi (Richard), Zhang
https://arxiv.org/abs/2510.08413

Prompts Generalize with Low Data: Non-vacuous Generalization Bounds for Optimizing Prompts with More Informative Priors
Many prompt engineering techniques have been successful in practice, even when optimizing over a large prompt space with with a small amount of task-specific data. Recent work has partially explained this success by showing generalization bounds which apply PAC-Bayes theory to the discrete prompt space, but they are non-vacuous only in data-rich scenarios. We argue that such widespread success can be more fully explained through more carefully considering data- or distribution-dependent perplex…

📥 Powerful import system directly in prompts: @./src/api.ts for single files, @./src/**/*.ts for glob patterns, @./file.ts:10-50 for line ranges, @./types.ts#UserInterface for TypeScript symbol extraction, @https://url for remote content
🎯 Template variables for dynamic prompts: {{ _stdin }} captures piped input, {{ _1 }} {{ _2 }} for positional arguments, {{ _args }} for numbered list of all args, custom variables like {{ _feature_name }} with CLI overrides --_feature_name "value…

SNAP: Towards Segmenting Anything in Any Point Cloud
Aniket Gupta, Hanhui Wang, Charles Saunders, Aruni RoyChowdhury, Hanumant Singh, Huaizu Jiang
https://arxiv.org/abs/2510.11565

SNAP: Towards Segmenting Anything in Any Point Cloud
Interactive 3D point cloud segmentation enables efficient annotation of complex 3D scenes through user-guided prompts. However, current approaches are typically restricted in scope to a single domain (indoor or outdoor), and to a single form of user interaction (either spatial clicks or textual prompts). Moreover, training on multiple datasets often leads to negative transfer, resulting in domain-specific tools that lack generalizability. To address these limitations, we present \textbf{SNAP} (…

From Prompts to Packets: A View from the Network on ChatGPT, Copilot, and Gemini
Antonio Montieri, Alfredo Nascita, Antonio Pescap\`e
https://arxiv.org/abs/2510.11269 https://…

From Prompts to Packets: A View from the Network on ChatGPT, Copilot, and Gemini
Generative AI (GenAI) chatbots are now pervasive in digital ecosystems, yet their network traffic remains largely underexplored. This study presents an in-depth investigation of traffic generated by three leading chatbots (ChatGPT, Copilot, and Gemini) when accessed via Android mobile apps for both text and image generation. Using a dedicated capture architecture, we collect and label two complementary workloads: a 60-hour generic dataset with unconstrained prompts, and a controlled dataset bui…

Do You Hear What I Mean? Quantifying the Instruction-Perception Gap in Instruction-Guided Expressive Text-To-Speech Systems
Yi-Cheng Lin, Huang-Cheng Chou, Tzu-Chieh Wei, Kuan-Yu Chen, Hung-yi Lee
https://arxiv.org/abs/2509.13989

Do You Hear What I Mean? Quantifying the Instruction-Perception Gap in Instruction-Guided Expressive Text-To-Speech Systems
Instruction-guided text-to-speech (ITTS) enables users to control speech generation through natural language prompts, offering a more intuitive interface than traditional TTS. However, the alignment between user style instructions and listener perception remains largely unexplored. This work first presents a perceptual analysis of ITTS controllability across two expressive dimensions (adverbs of degree and graded emotion intensity) and collects human ratings on speaker age and word-level emphas…

iCodeReviewer: Improving Secure Code Review with Mixture of Prompts
Yun Peng, Kisub Kim, Linghan Meng, Kui Liu
https://arxiv.org/abs/2510.12186 https://arx…

iCodeReviewer: Improving Secure Code Review with Mixture of Prompts
Code review is an essential process to ensure the quality of software that identifies potential software issues at an early stage of software development. Among all software issues, security issues are the most important to identify, as they can easily lead to severe software crashes and service disruptions. Recent research efforts have been devoted to automated approaches to reduce the manual efforts required in the secure code review process. Despite the progress, current automated approaches…

In the run-up to Dutch elections (29 October), it turns out that two Dutch MP's of Wilders' far-right PVV are using Sora AI to generate fake pictures of attractive blond women threatened by migrants. The prompts they used were published in this article.

Vanuit de PVV-fractie worden in het geheim met AI racistische ideeën gepromoot
Op sociale media delen PVV-fanaccounts levensechte AI-beelden, veelal van knappe, blonde vrouwen die worden bedreigd door migranten. Deze pagina’s blijken te worden beheerd door minstens twee PVV-Kamerleden.

Good luck to this campaign.
One of my best cycle tours followed the Wye valley to the sea at Chepstow.
It is a scandal what's been done to this rural watercourse.
River Wye pollution prompts UK's largest environmental lawsuit - BBC News
https://www.bbc.co.uk/news/articles/crrj4q

River Wye pollution prompts UK's largest environmental lawsuit
Livestock and water companies are accused of “extensive” pollution in the Wye, Lugg and Usk rivers.

Anthropic-Philosophin erklärt: So schreiben sie wirklich gute Prompts
Laut Amanda Askell von Anthropic ist Prompt Engineering mehr als nur Ausprobieren. Philosophie habe ihr geholfen, Anweisungen an KI-Tools klarer zu formulieren.

#TIL we were supposed to shout at AI.
https://www.robert-glaser.de/prompts-as-programs-in-gpt-5/

The End of Shouting: Prompts as Programs in GPT-5
Why does a prompt that works brilliantly with Claude suddenly break on GPT-5? And since when do we have to write prompts like rule-based programs? How do we enforce “creativity” when Marketing pushes for it? The answer is a fundamental paradigm shift: GPT-5 with reasoning (GPT-5 Thinking) follows instructions with

OpenAI says GPT‑5 instant and GPT‑5 thinking cut political bias by 30% from earlier models, and show greater robustness to charged prompts (Ashley Gold/Axios)
https://www.axios.com/2025/10/09/openai-gpt-5-least-biased-model

OpenAI says GPT-5 is its least biased model yet
New findings from OpenAI researchers found that "charged" prompts produced the most biased results.

No MoCap Needed: Post-Training Motion Diffusion Models with Reinforcement Learning using Only Textual Prompts
Girolamo Macaluso, Lorenzo Mandelli, Mirko Bicchierai, Stefano Berretti, Andrew D. Bagdanov
https://arxiv.org/abs/2510.06988

No MoCap Needed: Post-Training Motion Diffusion Models with Reinforcement Learning using Only Textual Prompts
Diffusion models have recently advanced human motion generation, producing realistic and diverse animations from textual prompts. However, adapting these models to unseen actions or styles typically requires additional motion capture data and full retraining, which is costly and difficult to scale. We propose a post-training framework based on Reinforcement Learning that fine-tunes pretrained motion diffusion models using only textual prompts, without requiring any motion ground truth. Our appr…

So Coupang didn't know it had been breached until one of its customers reported it to them.
This customer, surnamed Park, was informed of the breach by a stranger via email, who showed Park's 15 most recent emails with the retailer.
https://<…

A customer reported a suspicious email to Coupang. Then the platform announced a breach.
A suspicious email with the subject line “Your personal information may have been leaked” pushed Park, a software developer, into a decision he still describes as surreal: he reported it to Coupang.

Beyond Single-Granularity Prompts: A Multi-Scale Chain-of-Thought Prompt Learning for Graph
Ziyu Zheng, Yaming Yang, Ziyu Guan, Wei Zhao, Xinyan Huang, Weigang Lu
https://arxiv.org/abs/2510.09394

Beyond Single-Granularity Prompts: A Multi-Scale Chain-of-Thought Prompt Learning for Graph
The "pre-train, prompt'' paradigm, designed to bridge the gap between pre-training tasks and downstream objectives, has been extended from the NLP domain to the graph domain and has achieved remarkable progress. Current mainstream graph prompt-tuning methods modify input or output features using learnable prompt vectors. However, existing approaches are confined to single-granularity (e.g., node-level or subgraph-level) during prompt generation, overlooking the inherently multi-scale structural…

Keep Calm and Avoid Harmful Content: Concept Alignment and Latent Manipulation Towards Safer Answers
Ruben Belo, Claudia Soares, Marta Guimaraes
https://arxiv.org/abs/2510.12672

Keep Calm and Avoid Harmful Content: Concept Alignment and Latent Manipulation Towards Safer Answers
Large Language Models are susceptible to jailbreak attacks that bypass built-in safety guardrails (e.g., by tricking the model with adversarial prompts). We propose Concept Alignment and Concept Manipulation \textbf{CALM}, an inference-time method that suppresses harmful concepts by modifying latent representations of the last layer of the model, without retraining. Leveraging \gls*{cw} technique from Computer Vision combined with orthogonal projection, CALM removes unwanted latent directions a…

If you want to spend time on AI you can best spend it on lectures like this. No hype, just science, but in this case also very practical.
#AI

Stanford CS230 | Autumn 2025 | Lecture 7: Agents, Prompts, and RAG.
For more information about Stanford’s Artificial Intelligence professional and graduate programs, visit: https://stanford.io/aiNovember 11, 2025This lecture ...

🍮 Wissen zum Nachtisch: 🍨
Immer mehr Menschen sehen sich im beruflichen Umfeld genötigt, mit generativer #KI zu arbeiten.
Besonders Großunternehmen „überrollen“ damit ihre #Mitarbeiter. Es wird eine Art

Generative KI für Einsteiger - Praxisnahe Prompts für Privat und Beruf – einfach erklärt
Buchempfehlungen zu Klimawandel, Nachhaltigkeit, Umweltschutz und Klimaanpassung

Fei-Fei Li's World Labs launches Marble, its first world model to let users turn prompts, photos, and other media into editable 3D environments, after a beta (Rebecca Bellan/TechCrunch)
https://techcrunch.com/2025/11/12/fei-

Fei-Fei Li's World Labs speeds up the world model race with Marble, its first commercial product | TechCrunch
Marble is different from competitors like Odyssey, Decart, and Google's Genie because it creates persistent, downloadable 3D environments rather than generating worlds on-the-fly as you explore.

VLM-Guided Adaptive Negative Prompting for Creative Generation
Shelly Golan, Yotam Nitzan, Zongze Wu, Or Patashnik
https://arxiv.org/abs/2510.10715 https://

VLM-Guided Adaptive Negative Prompting for Creative Generation
Creative generation is the synthesis of new, surprising, and valuable samples that reflect user intent yet cannot be envisioned in advance. This task aims to extend human imagination, enabling the discovery of visual concepts that exist in the unexplored spaces between familiar domains. While text-to-image diffusion models excel at rendering photorealistic scenes that faithfully match user prompts, they still struggle to generate genuinely novel content. Existing approaches to enhance generativ…

If, like me, you weren’t familiar with #DSPy, this was a great talk given at the Databricks Data & AI summit to walk you through why you should care.
If you’re writing prompts by hand in your apps, stop now and read this article by @…

Let the Model Write the Prompt
Notes from a talk I delivered at the 2025 Data + AI Summit, detailing the problem with prompts in your code and how DSPy can make everything better.

Analyzing and Internalizing Complex Policy Documents for LLM Agents
Jiateng Liu, Zhenhailong Wang, Xiaojiang Huang, Yingjie Li, Xing Fan, Xiang Li, Chenlei Guo, Ruhi Sarikaya, Heng Ji
https://arxiv.org/abs/2510.11588

Analyzing and Internalizing Complex Policy Documents for LLM Agents
Large Language Model (LLM)-based agentic systems rely on in-context policy documents encoding diverse business rules. As requirements grow, these documents expand rapidly, causing high computational overhead. This motivates developing internalization methods that embed policy documents into model priors while preserving performance. Prior prompt compression work targets generic prompts, but agentic policy documents span multiple complexity levels and require deeper reasoning, making internaliza…

Learning to Rewrite Prompts for Bootstrapping LLMs on Downstream Tasks
Qinhao Zhou, Xiang Xiang, Kun He, John E. Hopcroft
https://arxiv.org/abs/2510.06695 https://

Learning to Rewrite Prompts for Bootstrapping LLMs on Downstream Tasks
In recent years, the growing interest in Large Language Models (LLMs) has significantly advanced prompt engineering, transitioning from manual design to model-based optimization. Prompts for LLMs generally comprise two components: the \textit{instruction}, which defines the task or objective, and the \textit{input}, which is tailored to the instruction type. In natural language generation (NLG) tasks such as machine translation, the \textit{input} component is particularly critical, while the \…

Flamed-TTS: Flow Matching Attention-Free Models for Efficient Generating and Dynamic Pacing Zero-shot Text-to-Speech
Hieu-Nghia Huynh-Nguyen, Huynh Nguyen Dang, Ngoc-Son Nguyen, Van Nguyen
https://arxiv.org/abs/2510.02848

Flamed-TTS: Flow Matching Attention-Free Models for Efficient Generating and Dynamic Pacing Zero-shot Text-to-Speech
Zero-shot Text-to-Speech (TTS) has recently advanced significantly, enabling models to synthesize speech from text using short, limited-context prompts. These prompts serve as voice exemplars, allowing the model to mimic speaker identity, prosody, and other traits without extensive speaker-specific data. Although recent approaches incorporating language models, diffusion, and flow matching have proven their effectiveness in zero-shot TTS, they still encounter challenges such as unreliable synth…

SRUM: Fine-Grained Self-Rewarding for Unified Multimodal Models
Weiyang Jin, Yuwei Niu, Jiaqi Liao, Chengqi Duan, Aoxue Li, Shenghua Gao, Xihui Liu
https://arxiv.org/abs/2510.12784

SRUM: Fine-Grained Self-Rewarding for Unified Multimodal Models
Recently, remarkable progress has been made in Unified Multimodal Models (UMMs), which integrate vision-language generation and understanding capabilities within a single framework. However, a significant gap exists where a model's strong visual understanding often fails to transfer to its visual generation. A model might correctly understand an image based on user instructions, yet be unable to generate a faithful image from text prompts. This phenomenon directly raises a compelling question: …

🎲 TextBandit: Evaluating Probabilistic Reasoning in LLMs Through Language-Only Decision Tasks
#llm

TextBandit: Evaluating Probabilistic Reasoning in LLMs Through Language-Only Decision Tasks
Large language models (LLMs) have shown to be increasingly capable of performing reasoning tasks, but their ability to make sequential decisions under uncertainty only using natural language remains underexplored. We introduce a novel benchmark in which LLMs interact with multi-armed bandit environments using purely textual feedback, "you earned a token", without access to numerical cues or explicit probabilities, resulting in the model to infer latent reward structures purely off linguistic cu…

🛠️ Monitor logs & events: user prompts, tool results with accept/reject decisions, #API requests with detailed cost/duration/token counts, error tracking with retry attempts & tool decision patterns
⚙️ Setup options: Launch via #VSCode or terminal with environment variables, configure…

When to Reason: Semantic Router for vLLM
Chen Wang, Xunzhuo Liu, Yuhan Liu, Yue Zhu, Xiangxi Mo, Junchen Jiang, Huamin Chen
https://arxiv.org/abs/2510.08731 https://

When to Reason: Semantic Router for vLLM
Large Language Models (LLMs) demonstrate substantial accuracy gains when augmented with reasoning modes such as chain-of-thought and inference-time scaling. However, reasoning also incurs significant costs in inference latency and token usage, with environmental and financial impacts, which are unnecessary for many simple prompts. We present a semantic router that classifies queries based on their reasoning requirements and selectively applies reasoning only when beneficial. Our approach achiev…

At least I got the day job to implement a "no AI browsers on work machines" policy.
If you CISO has not released a policy forbidding AI browsers at your company they are asleep at the wheel.
https://mstdn.social/@Hackread/115448122449051028

Runway launches Gen 4.5, a new text-to-video AI model that produces HD videos from written prompts and excels at physics; Gen 4.5 tops Video Arena's leaderboard (Ashley Capoot/CNBC)
https://www.cnbc.com/2025/12/01/runway-gen-4-5-video-model-google-open-ai.h…

Exposing Citation Vulnerabilities in Generative Engines
Riku Mochizuki, Shusuke Komatsu, Souta Noguchi, Kazuto Ataka
https://arxiv.org/abs/2510.06823 https://

Exposing Citation Vulnerabilities in Generative Engines
We analyze answers generated by generative engines (GEs) from the perspectives of citation publishers and the content-injection barrier, defined as the difficulty for attackers to manipulate answers to user prompts by placing malicious content on the web. GEs integrate two functions: web search and answer generation that cites web pages using large language models. Because anyone can publish information on the web, GEs are vulnerable to poisoning attacks. Existing studies of citation evaluation…

Gimme a shred of what's being sunk into LLMs and I'll increase the number of humanities and arts scholars by orders of magnitude, fund them for all eternity, and avoid wrecking the planet (setting aside discursive analysis, occasionally impenetrable buzzwords, and the newest form of "modern" (postmodern?) artistic expression). Give us your prompts (and several years) and we'll answer them. So what if we'd sometimes be wrong? At least the many more who could live fro…

Two Agents, One Prompt, and Your Weight
Elchanan Mossel, Amnon Schrieber
https://arxiv.org/abs/2510.08232 https://arxiv.org/pdf/2510.08232

Two Agents, One Prompt, and Your Weight
We investigate a quantitative variant of the classic Two Doors logic puzzle, in which the answer space is no longer binary, for example when the goal is to recover a numerical fact (such as one's true weight) rather than choose between two doors. The puzzle retains the original structure: one agent always tells the truth, the other always lies. Our central contribution is to identify a class of self-referential prompts that successfully extract the correct quantitative answer under minimal assu…

Egal zu welchem Thema, es wird immer unterschiedliche Positionen von Personen geben, die zu diesem Thema als Experte gelten. In diesen widersprüchlichen Meinungen öffnet sich ein Raum, in dem eine eigene Stimme entstehen kann:
"Nothing prepared me better for adult intellectual life than getting two sets of contradictory feedback on every essay I wrote for those classes, because I had to decide, over and over, what to make of these expressions of my teachers’ authority. Yes, they k…

What Happened When I Tried to Replace Myself with ChatGPT in My English Classroom
My students call it “Chat,” a cute nickname they all seem to have agreed on at some point. They use it to make study guides, interpret essay prompts, and register for classes, turning it loose on t…

“Alliances are built on common values and a common threat perception,”
said Danish Defense Analyst Jacob Kaarsbo
“Trump shares neither of those with us
and I would argue he doesn’t share it with most Europeans.”
https://www.

Google rolls out Nano Banana image editing upgrades in Photos, including a "help me edit" feature that lets users make edits using text or voice prompts (Elyse Betters Picaro/ZDNET)
https://www.zdnet.com/article/your-google-

Your Google Photos just got 3 huge Nano Banana image editing upgrades - for free
Google Photos just made it easier than ever to create realistic AI edits, thanks to Google DeepMind's top image-editing model.

SafeMT: Multi-turn Safety for Multimodal Language Models
Han Zhu, Juntao Dai, Jiaming Ji, Haoran Li, Chengkun Cai, Pengcheng Wen, Chi-Min Chan, Boyuan Chen, Yaodong Yang, Sirui Han, Yike Guo
https://arxiv.org/abs/2510.12133

SafeMT: Multi-turn Safety for Multimodal Language Models
With the widespread use of multi-modal Large Language models (MLLMs), safety issues have become a growing concern. Multi-turn dialogues, which are more common in everyday interactions, pose a greater risk than single prompts; however, existing benchmarks do not adequately consider this situation. To encourage the community to focus on the safety issues of these models in multi-turn dialogues, we introduce SafeMT, a benchmark that features dialogues of varying lengths generated from harmful quer…

🍮 Wissen zum Nachtisch: 🍨
Immer mehr Menschen sehen sich im beruflichen Umfeld genötigt, mit generativer #KI zu arbeiten.
Besonders Großunternehmen „überrollen“ damit ihre #Mitarbeiter. Es wird eine Art

Generative KI für Einsteiger - Praxisnahe Prompts für Privat und Beruf – einfach erklärt
Buchempfehlungen zu Klimawandel, Nachhaltigkeit, Umweltschutz und Klimaanpassung

Not to imply any disrespect, just struggling with an old brain vs the collapse of Internet Search, but I can't shake this sense of recalling that, long ago, #JaneGoodall shot someone, a poacher or such, can't recall exactly, maybe it wasnt even Jane, but Search refuses to understand ant permutations of my prompts.

X2Video: Adapting Diffusion Models for Multimodal Controllable Neural Video Rendering
Zhitong Huang, Mohan Zhang, Renhan Wang, Rui Tang, Hao Zhu, Jing Liao
https://arxiv.org/abs/2510.08530

X2Video: Adapting Diffusion Models for Multimodal Controllable Neural Video Rendering
We present X2Video, the first diffusion model for rendering photorealistic videos guided by intrinsic channels including albedo, normal, roughness, metallicity, and irradiance, while supporting intuitive multi-modal controls with reference images and text prompts for both global and local regions. The intrinsic guidance allows accurate manipulation of color, material, geometry, and lighting, while reference images and text prompts provide intuitive adjustments in the absence of intrinsic inform…

TIT-Score: Evaluating Long-Prompt Based Text-to-Image Alignment via Text-to-Image-to-Text Consistency
Juntong Wang, Huiyu Duan, Jiarui Wang, Ziheng Jia, Guangtao Zhai, Xiongkuo Min
https://arxiv.org/abs/2510.02987

TIT-Score: Evaluating Long-Prompt Based Text-to-Image Alignment via Text-to-Image-to-Text Consistency
With the rapid advancement of large multimodal models (LMMs), recent text-to-image (T2I) models can generate high-quality images and demonstrate great alignment to short prompts. However, they still struggle to effectively understand and follow long and detailed prompts, displaying inconsistent generation. To address this challenge, we introduce LPG-Bench, a comprehensive benchmark for evaluating long-prompt-based text-to-image generation. LPG-Bench features 200 meticulously crafted prompts wit…

Leaked Meta internal document shows updated training protocols for its AI chatbot to refuse any prompts involving sexual roleplay with minors (Jyoti Mann/Business Insider)
https://www.businessinsider.com/meta-ai-chatbot-children-ftc-safety-guidelines-202…

Leaked Meta guidelines show how it trains AI chatbots to respond to child sexual exploitation prompts
A leaked Meta document shows how its AI chatbot is trained to handle child sexual exploitation prompts as it faces FTC scrutiny.

Prompts as Software Engineering Artifacts: A Research Agenda and Preliminary Findings
Hugo Villamizar, Jannik Fischbach, Alexander Korn, Andreas Vogelsang, Daniel Mendez
https://arxiv.org/abs/2509.17548

Prompts as Software Engineering Artifacts: A Research Agenda and Preliminary Findings
Developers now routinely interact with large language models (LLMs) to support a range of software engineering (SE) tasks. This prominent role positions prompts as potential SE artifacts that, like other artifacts, may require systematic development, documentation, and maintenance. However, little is known about how prompts are actually used and managed in LLM-integrated workflows, what challenges practitioners face, and whether the benefits of systematic prompt management outweigh the associat…

Efficient Prompt Optimisation for Legal Text Classification with Proxy Prompt Evaluator
Hyunji Lee, Kevin Chenhao Li, Matthias Grabmair, Shanshan Xu
https://arxiv.org/abs/2510.08524

Efficient Prompt Optimisation for Legal Text Classification with Proxy Prompt Evaluator
Prompt optimization aims to systematically refine prompts to enhance a language model's performance on specific tasks. Fairness detection in Terms of Service (ToS) clauses is a challenging legal NLP task that demands carefully crafted prompts to ensure reliable results. However, existing prompt optimization methods are often computationally expensive due to inefficient search strategies and costly prompt candidate scoring. In this paper, we propose a framework that combines Monte Carlo Tree Sea…

🍮 Wissen zum Nachtisch: 🍨
Immer mehr Menschen sehen sich im beruflichen Umfeld genötigt, mit generativer #KI zu arbeiten.
Besonders Großunternehmen „überrollen“ damit ihre #Mitarbeiter. Es wird eine Art

Generative KI für Einsteiger - Praxisnahe Prompts für Privat und Beruf – einfach erklärt
Buchempfehlungen zu Klimawandel, Nachhaltigkeit, Umweltschutz und Klimaanpassung

SafeGuider: Robust and Practical Content Safety Control for Text-to-Image Models
Peigui Qi, Kunsheng Tang, Wenbo Zhou, Weiming Zhang, Nenghai Yu, Tianwei Zhang, Qing Guo, Jie Zhang
https://arxiv.org/abs/2510.05173

SafeGuider: Robust and Practical Content Safety Control for Text-to-Image Models
Text-to-image models have shown remarkable capabilities in generating high-quality images from natural language descriptions. However, these models are highly vulnerable to adversarial prompts, which can bypass safety measures and produce harmful content. Despite various defensive strategies, achieving robustness against attacks while maintaining practical utility in real-world applications remains a significant challenge. To address this issue, we first conduct an empirical study of the text e…

Mono4DEditor: Text-Driven 4D Scene Editing from Monocular Video via Point-Level Localization of Language-Embedded Gaussians
Jin-Chuan Shi, Chengye Su, Jiajun Wang, Ariel Shamir, Miao Wang
https://arxiv.org/abs/2510.09438

Mono4DEditor: Text-Driven 4D Scene Editing from Monocular Video via Point-Level Localization of Language-Embedded Gaussians
Editing 4D scenes reconstructed from monocular videos based on text prompts is a valuable yet challenging task with broad applications in content creation and virtual environments. The key difficulty lies in achieving semantically precise edits in localized regions of complex, dynamic scenes, while preserving the integrity of unedited content. To address this, we introduce Mono4DEditor, a novel framework for flexible and accurate text-driven 4D scene editing. Our method augments 3D Gaussians wi…

Beyond Western Politics: Cross-Cultural Benchmarks for Evaluating Partisan Associations in LLMs
Divyanshu Kumar, Ishita Gupta, Nitin Aravind Birur, Tanay Baswa, Sahil Agarwal, Prashanth Harshangi
https://arxiv.org/abs/2509.22711

Beyond Western Politics: Cross-Cultural Benchmarks for Evaluating Partisan Associations in LLMs
Partisan bias in LLMs has been evaluated to assess political leanings, typically through a broad lens and largely in Western contexts. We move beyond identifying general leanings to examine harmful, adversarial representational associations around political leaders and parties. To do so, we create datasets \textit{NeutQA-440} (non-adversarial prompts) and \textit{AdverQA-440} (adversarial prompts), which probe models for comparative plausibility judgments across the USA and India. Results show …

The indictment of former F.B.I. director, James B. Comey
was just two pages long and contained so little detail that it was hard for legal experts to assess its merits.
Some former government lawyers, as well as prominent Democrats, said the case appeared so flimsy that it was likely to fail.
Some Republicans also expressed concerns

Brief Comey Indictment Prompts Questions and Criticism
The two-page indictment of James B. Comey, the former F.B.I. director, offered little indication of the evidence that would be presented at a trial. Some legal experts called it flimsy and dangerous.

Two Modes of Reflection: How Temporal, Spatial, and Social Distances Affect Reflective Writing in Family Caregiving
Shunpei Norihama, Yuka Iwane, Jo Takezawa, Simo Hosio, Mari Hirano, Naomi Yamashita, Koji Yatani
https://arxiv.org/abs/2510.05510

Two Modes of Reflection: How Temporal, Spatial, and Social Distances Affect Reflective Writing in Family Caregiving
Writing about personal experiences can improve well-being, but for family caregivers, fixed or user-initiated schedules often miss the right moments. Drawing on Construal Level Theory, we conducted a three-week field study with 47 caregivers using a chatbot that delivered daily reflective writing prompts and captured temporal, spatial, and social contexts. We collected 958 writing entries, resulting in 5,412 coded segments. Our Analysis revealed two reflective modes. Under proximal conditions, …

Runway launches Gen 4.5, a new text-to-video AI model that produces HD videos from written prompts and excels at physics; Gen 4.5 tops Video Arena's leaderboard (Ashley Capoot/CNBC)
https://www.cnbc.com/2025/12/01/runway-gen-4-5-video-model-google-open-ai.h…

Runway rolls out new AI video model that beats Google, OpenAI in key benchmark
Gen 4.5, a new AI model that allows users to generate high-definition videos based on written prompts.

SemShareKV: Efficient KVCache Sharing for Semantically Similar Prompts via Token-Level LSH Matching
Xinye Zhao, Spyridon Mastorakis
https://arxiv.org/abs/2509.24832 https://

SemShareKV: Efficient KVCache Sharing for Semantically Similar Prompts via Token-Level LSH Matching
As large language models (LLMs) continue to scale, the memory footprint of key-value (KV) caches during inference has become a significant bottleneck. Existing approaches primarily focus on compressing KV caches within a single prompt or reusing shared prefixes or frequently ocurred text segments across prompts. However, such strategies are limited in scenarios where prompts are semantically similar but lexically different, which frequently occurs in tasks such as multi-document summarization a…

MetaLogic: Robustness Evaluation of Text-to-Image Models via Logically Equivalent Prompts
Yifan Shen, Yangyang Shu, Hye-young Paik, Yulei Sui
https://arxiv.org/abs/2510.00796 ht…

MetaLogic: Robustness Evaluation of Text-to-Image Models via Logically Equivalent Prompts
Recent advances in text-to-image (T2I) models, especially diffusion-based architectures, have significantly improved the visual quality of generated images. However, these models continue to struggle with a critical limitation: maintaining semantic consistency when input prompts undergo minor linguistic variations. Despite being logically equivalent, such prompt pairs often yield misaligned or semantically inconsistent images, exposing a lack of robustness in reasoning and generalisation. To ad…

Rule Encoding and Compliance in Large Language Models: An Information-Theoretic Analysis
Joachim Diederich
https://arxiv.org/abs/2510.05106 https://arxiv.o…

Rule Encoding and Compliance in Large Language Models: An Information-Theoretic Analysis
The design of safety-critical agents based on large language models (LLMs) requires more than simple prompt engineering. This paper presents a comprehensive information-theoretic analysis of how rule encodings in system prompts influence attention mechanisms and compliance behaviour. We demonstrate that rule formats with low syntactic entropy and highly concentrated anchors reduce attention entropy and improve pointer fidelity, but reveal a fundamental trade-off between anchor redundancy and at…

Reinforce-Ada: An Adaptive Sampling Framework for Reinforce-Style LLM Training
Wei Xiong, Chenlu Ye, Baohao Liao, Hanze Dong, Xinxing Xu, Christof Monz, Jiang Bian, Nan Jiang, Tong Zhang
https://arxiv.org/abs/2510.04996

Reinforce-Ada: An Adaptive Sampling Framework for Reinforce-Style LLM Training
Reinforcement learning applied to large language models (LLMs) for reasoning tasks is often bottlenecked by unstable gradient estimates due to fixed and uniform sampling of responses across prompts. Prior work such as GVM-RAFT addresses this by dynamically allocating inference budget per prompt to minimize stochastic gradient variance under a budget constraint. Inspired by this insight, we propose Reinforce-Ada, an adaptive sampling framework for online RL post-training of LLMs that continuousl…

Reading Between the Lines: Scalable User Feedback via Implicit Sentiment in Developer Prompts
Daye Nam, Malgorzata Salawa, Satish Chandra
https://arxiv.org/abs/2509.18361 https:…

The Limits of Goal-Setting Theory in LLM-Driven Assessment
Mrityunjay Kumar
https://arxiv.org/abs/2510.06997 https://arxiv.org/pdf/2510.06997

The Limits of Goal-Setting Theory in LLM-Driven Assessment
Many users interact with AI tools like ChatGPT using a mental model that treats the system as human-like, which we call Model H. According to goal-setting theory, increased specificity in goals should reduce performance variance. If Model H holds, then prompting a chatbot with more detailed instructions should lead to more consistent evaluation behavior. This paper tests that assumption through a controlled experiment in which ChatGPT evaluated 29 student submissions using four prompts with i…

The meaning of prompts and the prompts of meaning: Semiotic reflections and modelling
Martin Thellefsen, Amalia Nurma Dewi, Bent Sorensen
https://arxiv.org/abs/2509.14250 https:…

The meaning of prompts and the prompts of meaning: Semiotic reflections and modelling
This paper explores prompts and prompting in large language models (LLMs) as dynamic semiotic phenomena, drawing on Peirce's triadic model of signs, his nine sign types, and the Dynacom model of communication. The aim is to reconceptualize prompting not as a technical input mechanism but as a communicative and epistemic act involving an iterative process of sign formation, interpretation, and refinement. The theoretical foundation rests on Peirce's semiotics, particularly the interplay between …

ObCLIP: Oblivious CLoud-Device Hybrid Image Generation with Privacy Preservation
Haoqi Wu, Wei Dai, Ming Xu, Li Wang, Qiang Yan
https://arxiv.org/abs/2510.04153 https://

ObCLIP: Oblivious CLoud-Device Hybrid Image Generation with Privacy Preservation
Diffusion Models have gained significant popularity due to their remarkable capabilities in image generation, albeit at the cost of intensive computation requirement. Meanwhile, despite their widespread deployment in inference services such as Midjourney, concerns about the potential leakage of sensitive information in uploaded user prompts have arisen. Existing solutions either lack rigorous privacy guarantees or fail to strike an effective balance between utility and efficiency. To bridge thi…

Multimodal Policy Internalization for Conversational Agents
Zhenhailong Wang, Jiateng Liu, Amin Fazel, Ritesh Sarkhel, Xing Fan, Xiang Li, Chenlei Guo, Heng Ji, Ruhi Sarikaya
https://arxiv.org/abs/2510.09474

Multimodal Policy Internalization for Conversational Agents
Modern conversational agents like ChatGPT and Alexa+ rely on predefined policies specifying metadata, response styles, and tool-usage rules. As these LLM-based systems expand to support diverse business and user queries, such policies, often implemented as in-context prompts, are becoming increasingly complex and lengthy, making faithful adherence difficult and imposing large fixed computational costs. With the rise of multimodal agents, policies that govern visual and multimodal behaviors are …

Researchers claim that prompts framed as riddle-like poems could skirt AI chatbots' safety features designed to block production of explicit or harmful content (Robert Hart/The Verge)
https://www.theverge.com/report/838167/ai-chatbots-can-be-wooed-…

AI chatbots can be wooed into crimes with poetry
New research suggests riddle-like poems are remarkably effective at circumventing AI safety features. 

Inoculation Prompting: Instructing LLMs to misbehave at train-time improves test-time alignment
Nevan Wichers, Aram Ebtekar, Ariana Azarbal, Victor Gillioz, Christine Ye, Emil Ryd, Neil Rathi, Henry Sleight, Alex Mallen, Fabien Roger, Samuel Marks
https://arxiv.org/abs/2510.05024

Inoculation Prompting: Instructing LLMs to misbehave at train-time improves test-time alignment
Large language models are sometimes trained with imperfect oversight signals, leading to undesired behaviors such as reward hacking and sycophancy. Improving oversight quality can be expensive or infeasible, motivating methods that improve learned behavior despite an imperfect training signal. We introduce Inoculation Prompting (IP), a simple but counterintuitive technique that prevents learning of an undesired behavior by modifying training prompts to explicitly request it. For example, to ino…

US-Ermittler haben erstmals eine #Durchsuchungsanordnung gegen #OpenAI erwirkt, um Nutzerdaten von #ChatGPT in einem Fall von

ChatGPT-Überwachung: OpenAI gibt Daten an US-Behörden weiter
Erstmals fordern US-Behörden Nutzerdaten wegen harmloser ChatGPT-Prompts an. Der Fall zeigt, wie generative KI zur Hilfestellung bei Verbrechen wird

Solving Math Word Problems Using Estimation Verification and Equation Generation
Mitchell Piehl, Dillon Wilson, Ananya Kalita, Jugal Kalita
https://arxiv.org/abs/2509.18565 http…

Research: US companies announced 153,074 job cuts in October, the most in 20 years and up 3x on October 2024, as AI reshapes industries and cost-cutting rises (Julia Fanzeres/Bloomberg)
https://www.bloomberg.com/news/articles/20

Unlocking Vision-Language Models for Video Anomaly Detection via Fine-Grained Prompting
Shu Zou, Xinyu Tian, Lukas Wesemann, Fabian Waschkowski, Zhaoyuan Yang, Jing Zhang
https://arxiv.org/abs/2510.02155

Unlocking Vision-Language Models for Video Anomaly Detection via Fine-Grained Prompting
Prompting has emerged as a practical way to adapt frozen vision-language models (VLMs) for video anomaly detection (VAD). Yet, existing prompts are often overly abstract, overlooking the fine-grained human-object interactions or action semantics that define complex anomalies in surveillance videos. We propose ASK-Hint, a structured prompting framework that leverages action-centric knowledge to elicit more accurate and interpretable reasoning from frozen VLMs. Our approach organizes prompts into…

Wabi, a "YouTube for apps" launched by Replika founder Eugenia Kuyda to let users create and share mini apps without coding, raised a $20M pre-seed (TechCrunch)
https://techcrunch.com/2025/11/05/replika-founder-raises-20m-pre-seed…

Replika founder raises $20M pre-seed for Wabi, the ‘YouTube of apps’  | TechCrunch
Wabi, a startup from the founder of Replika, has just raised a $20 million pre-seed round. Wabi is like "YouTube for apps" -- a social platform where anyone can use prompts to instantly create mini apps and share them with friends.

DecipherGuard: Understanding and Deciphering Jailbreak Prompts for a Safer Deployment of Intelligent Software Systems
Rui Yang, Michael Fu, Chakkrit Tantithamthavorn, Chetan Arora, Gunel Gulmammadova, Joey Chua
https://arxiv.org/abs/2509.16870

DecipherGuard: Understanding and Deciphering Jailbreak Prompts for a Safer Deployment of Intelligent Software Systems
Intelligent software systems powered by Large Language Models (LLMs) are increasingly deployed in critical sectors, raising concerns about their safety during runtime. Through an industry-academic collaboration when deploying an LLM-powered virtual customer assistant, a critical software engineering challenge emerged: how to enhance a safer deployment of LLM-powered software systems at runtime? While LlamaGuard, the current state-of-the-art runtime guardrail, offers protection against unsafe in…

GenPilot: A Multi-Agent System for Test-Time Prompt Optimization in Image Generation
Wen Ye, Zhaocheng Liu, Yuwei Gui, Tingyu Yuan, Yunyue Su, Bowen Fang, Chaoyang Zhao, Qiang Liu, Liang Wang
https://arxiv.org/abs/2510.07217

GenPilot: A Multi-Agent System for Test-Time Prompt Optimization in Image Generation
Text-to-image synthesis has made remarkable progress, yet accurately interpreting complex and lengthy prompts remains challenging, often resulting in semantic inconsistencies and missing details. Existing solutions, such as fine-tuning, are model-specific and require training, while prior automatic prompt optimization (APO) approaches typically lack systematic error analysis and refinement strategies, resulting in limited reliability and effectiveness. Meanwhile, test-time scaling methods opera…

NLP Methods for Detecting Novel LLM Jailbreaks and Keyword Analysis with BERT
John Hawkins, Aditya Pramar, Rodney Beard, Rohitash Chandra
https://arxiv.org/abs/2510.01644 https:…

NLP Methods for Detecting Novel LLM Jailbreaks and Keyword Analysis with BERT
Large Language Models (LLMs) suffer from a range of vulnerabilities that allow malicious users to solicit undesirable responses through manipulation of the input text. These so-called jailbreak prompts are designed to trick the LLM into circumventing the safety guardrails put in place to keep responses acceptable to the developer's policies. In this study, we analyse the ability of different machine learning models to distinguish jailbreak prompts from genuine uses, including looking at our abi…

StyleKeeper: Prevent Content Leakage using Negative Visual Query Guidance
Jaeseok Jeong, Junho Kim, Gayoung Lee, Yunjey Choi, Youngjung Uh
https://arxiv.org/abs/2510.06827 https…

StyleKeeper: Prevent Content Leakage using Negative Visual Query Guidance
In the domain of text-to-image generation, diffusion models have emerged as powerful tools. Recently, studies on visual prompting, where images are used as prompts, have enabled more precise control over style and content. However, existing methods often suffer from content leakage, where undesired elements of the visual style prompt are transferred along with the intended style. To address this issue, we 1) extend classifier-free guidance (CFG) to utilize swapping self-attention and propose 2)…

AutoRed: A Free-form Adversarial Prompt Generation Framework for Automated Red Teaming
Muxi Diao, Yutao Mou, Keqing He, Hanbo Song, Lulu Zhao, Shikun Zhang, Wei Ye, Kongming Liang, Zhanyu Ma
https://arxiv.org/abs/2510.08329

AutoRed: A Free-form Adversarial Prompt Generation Framework for Automated Red Teaming
The safety of Large Language Models (LLMs) is crucial for the development of trustworthy AI applications. Existing red teaming methods often rely on seed instructions, which limits the semantic diversity of the synthesized adversarial prompts. We propose AutoRed, a free-form adversarial prompt generation framework that removes the need for seed instructions. AutoRed operates in two stages: (1) persona-guided adversarial instruction generation, and (2) a reflection loop to iteratively refine low…

The Alignment Waltz: Jointly Training Agents to Collaborate for Safety
Jingyu Zhang, Haozhu Wang, Eric Michael Smith, Sid Wang, Amr Sharaf, Mahesh Pasupuleti, Benjamin Van Durme, Daniel Khashabi, Jason Weston, Hongyuan Zhan
https://arxiv.org/abs/2510.08240

The Alignment Waltz: Jointly Training Agents to Collaborate for Safety
Harnessing the power of LLMs requires a delicate dance between being helpful and harmless. This creates a fundamental tension between two competing challenges: vulnerability to adversarial attacks that elicit unsafe content, and a tendency for overrefusal on benign but sensitive prompts. Current approaches often navigate this dance with safeguard models that completely reject any content that contains unsafe portions. This approach cuts the music entirely-it may exacerbate overrefusals and fail…

Beyond Over-Refusal: Scenario-Based Diagnostics and Post-Hoc Mitigation for Exaggerated Refusals in LLMs
Shuzhou Yuan, Ercong Nie, Yinuo Sun, Chenxuan Zhao, William LaCroix, Michael F\"arber
https://arxiv.org/abs/2510.08158

Beyond Over-Refusal: Scenario-Based Diagnostics and Post-Hoc Mitigation for Exaggerated Refusals in LLMs
Large language models (LLMs) frequently produce false refusals, declining benign requests that contain terms resembling unsafe queries. We address this challenge by introducing two comprehensive benchmarks: the Exaggerated Safety Benchmark (XSB) for single-turn prompts, annotated with "Focus" keywords that identify refusal-inducing triggers, and the Multi-turn Scenario-based Exaggerated Safety Benchmark (MS-XSB), which systematically evaluates refusal calibration in realistic, context-rich dial…

OpenAI launches new ChatGPT parental safety controls for users aged 13 to 18, including parental and law enforcement alerts for chats about self-harm or suicide (Reece Rogers/Wired)
https://www.wired.com/story/openai-teen-safety-tools-chatgpt-parents-s…

OpenAI Adds Parental Safety Controls for Teen ChatGPT Users. Here’s What to Expect
OpenAI’s review process for teenage ChatGPT users who are flagged for suicidal ideation includes human moderators. Parents can expect an alert about alarming prompts within hours.

ComposeMe: Attribute-Specific Image Prompts for Controllable Human Image Generation
Guocheng Gordon Qian, Daniil Ostashev, Egor Nemchinov, Avihay Assouline, Sergey Tulyakov, Kuan-Chieh Jackson Wang, Kfir Aberman
https://arxiv.org/abs/2509.18092

ComposeMe: Attribute-Specific Image Prompts for Controllable Human Image Generation
Generating high-fidelity images of humans with fine-grained control over attributes such as hairstyle and clothing remains a core challenge in personalized text-to-image synthesis. While prior methods emphasize identity preservation from a reference image, they lack modularity and fail to provide disentangled control over specific visual attributes. We introduce a new paradigm for attribute-specific image prompting, in which distinct sets of reference images are used to guide the generation of …

Adobe unveils Project Moonlight, an AI agent on its Firefly platform designed to act as a creative director for social media campaigns via text prompts (The Verge)
https://www.theverge.com/news/807457/adobe-ai-agent-project-moonlight

Adobe’s AI social media admin is here with ‘Project Moonlight’
Adobe announced Project Moonlight, an AI agent that uses other Adobe AI tools to create social media posts that match your tone and style.

Are LLMs Reliable Rankers? Rank Manipulation via Two-Stage Token Optimization
Tiancheng Xing, Jerry Li, Yixuan Du, Xiyang Hu
https://arxiv.org/abs/2510.06732 https://

Are LLMs Reliable Rankers? Rank Manipulation via Two-Stage Token Optimization
Large language models (LLMs) are increasingly used as rerankers in information retrieval, yet their ranking behavior can be steered by small, natural-sounding prompts. To expose this vulnerability, we present Rank Anything First (RAF), a two-stage token optimization method that crafts concise textual perturbations to consistently promote a target item in LLM-generated rankings while remaining hard to detect. Stage 1 uses Greedy Coordinate Gradient to shortlist candidate tokens at the current po…

Dynamic Prompt Generation for Interactive 3D Medical Image Segmentation Training
Tidiane Camaret Ndir, Alexander Pfefferle, Robin Tibor Schirrmeister
https://arxiv.org/abs/2510.03189

Dynamic Prompt Generation for Interactive 3D Medical Image Segmentation Training
Interactive 3D biomedical image segmentation requires efficient models that can iteratively refine predictions based on user prompts. Current foundation models either lack volumetric awareness or suffer from limited interactive capabilities. We propose a training strategy that combines dynamic volumetric prompt generation with content-aware adaptive cropping to optimize the use of the image encoder. Our method simulates realistic user interaction patterns during training while addressing the co…

Aligning Large Language Models via Fully Self-Synthetic Data
Shangjian Yin, Zhepei Wei, Xinyu Zhu, Wei-Lin Chen, Yu Meng
https://arxiv.org/abs/2510.06652 https://

Aligning Large Language Models via Fully Self-Synthetic Data
Traditional reinforcement learning from human feedback (RLHF) for large language models (LLMs) relies on expensive human-annotated datasets, while Reinforcement Learning from AI Feedback (RLAIF) also incurs significant costs, requiring the collection of diverse prompts and corresponding responses, often necessitating external reward models or proprietary models like GPT-4 to annotate preference pairs. In this work, we introduce Self-Alignment Optimization (SAO), a fully self-synthetic framework…

Sources: OpenAI is developing AI tools to generate music from text and audio prompts, including capabilities such as adding guitar accompaniment to vocal tracks (The Information)
https://www.theinformation.com/articles/openai-plots-gene…

OpenAI Plots Generating AI Music in Potential Rivalry With Startup Suno
OpenAI recently made a splash with artificial intelligence that generates short videos from text prompts—say, a scene from “Lord of the Rings” but with Gen Z characters, or Pokémon’s Pikachu as “The Godfather.” Now the company has set its sights on doing the same thing with music. OpenAI staff ...

Probing Social Identity Bias in Chinese LLMs with Gendered Pronouns and Social Groups
Geng Liu, Feng Li, Junjie Mu, Mengxiao Zhu, Francesco Pierri
https://arxiv.org/abs/2510.06974

Probing Social Identity Bias in Chinese LLMs with Gendered Pronouns and Social Groups
Large language models (LLMs) are increasingly deployed in user-facing applications, raising concerns about their potential to reflect and amplify social biases. We investigate social identity framing in Chinese LLMs using Mandarin-specific prompts across ten representative Chinese LLMs, evaluating responses to ingroup ("We") and outgroup ("They") framings, and extending the setting to 240 social groups salient in the Chinese context. To complement controlled experiments, we further analyze Chin…

When and Where do Events Switch in Multi-Event Video Generation?
Ruotong Liao, Guowen Huang, Qing Cheng, Thomas Seidl, Daniel Cremers, Volker Tresp
https://arxiv.org/abs/2510.03049

When and Where do Events Switch in Multi-Event Video Generation?
Text-to-video (T2V) generation has surged in response to challenging questions, especially when a long video must depict multiple sequential events with temporal coherence and controllable content. Existing methods that extend to multi-event generation omit an inspection of the intrinsic factor in event shifting. The paper aims to answer the central question: When and where multi-event prompts control event transition during T2V generation. This work introduces MEve, a self-curated prompt suite…

Privacy startup Duality says it has developed a private LLM inference framework that uses fully homomorphic encryption to let LLMs answer encrypted prompts (Rina Diane Caballar/IEEE Spectrum)
https://spectrum.ieee.org/homomorphic-encryption-llm

Homomorphic Encryption LLM Secures AI Chats
Can your chatbot keep secrets? Discover how Duality's tech is changing the privacy game for AI conversations.

Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)
Om Dobariya, Akhil Kumar
https://arxiv.org/abs/2510.04950 https://ar…

Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)
The wording of natural language prompts has been shown to influence the performance of large language models (LLMs), yet the role of politeness and tone remains underexplored. In this study, we investigate how varying levels of prompt politeness affect model accuracy on multiple-choice questions. We created a dataset of 50 base questions spanning mathematics, science, and history, each rewritten into five tone variants: Very Polite, Polite, Neutral, Rude, and Very Rude, yielding 250 unique prom…

Meta rolls out an AI assistant for Facebook Dating in the US and Canada, to reduce "swipe fatigue" by letting users find matches via natural language prompts (Sydney Bradley/Business Insider)
https://africa.businessinsider.com/new

Why Facebook thinks AI can help you land better dates
Facebook has an answer for swipe fatigue

TeachLM: Post-Training LLMs for Education Using Authentic Learning Data
Janos Perczel, Jin Chow, Dorottya Demszky
https://arxiv.org/abs/2510.05087 https://…

TeachLM: Post-Training LLMs for Education Using Authentic Learning Data
The promise of generative AI to revolutionize education is constrained by the pedagogical limits of large language models (LLMs). A major issue is the lack of access to high-quality training data that reflect the learning of actual students. Prompt engineering has emerged as a stopgap, but the ability of prompts to encode complex pedagogical strategies in rule-based natural language is inherently limited. To address this gap we introduce TeachLM - an LLM optimized for teaching through parameter…

SocialHarmBench: Revealing LLM Vulnerabilities to Socially Harmful Requests
Punya Syon Pandey, Hai Son Le, Devansh Bhardwaj, Rada Mihalcea, Zhijing Jin
https://arxiv.org/abs/2510.04891

SocialHarmBench: Revealing LLM Vulnerabilities to Socially Harmful Requests
Large language models (LLMs) are increasingly deployed in contexts where their failures can have direct sociopolitical consequences. Yet, existing safety benchmarks rarely test vulnerabilities in domains such as political manipulation, propaganda and disinformation generation, or surveillance and information control. We introduce SocialHarmBench, a dataset of 585 prompts spanning 7 sociopolitical categories and 34 countries, designed to surface where LLMs most acutely fail in politically charge…

Imperceptible Jailbreaking against Large Language Models
Kuofeng Gao, Yiming Li, Chao Du, Xin Wang, Xingjun Ma, Shu-Tao Xia, Tianyu Pang
https://arxiv.org/abs/2510.05025 https:/…

Imperceptible Jailbreaking against Large Language Models
Jailbreaking attacks on the vision modality typically rely on imperceptible adversarial perturbations, whereas attacks on the textual modality are generally assumed to require visible modifications (e.g., non-semantic suffixes). In this paper, we introduce imperceptible jailbreaks that exploit a class of Unicode characters called variation selectors. By appending invisible variation selectors to malicious questions, the jailbreak prompts appear visually identical to original malicious questions…