arxiv.orgResearch PaperTechnology & Ethics
Context - Authors
- Qiushi Han, Keya Hu, Linlu Qiu, Cathy Wu, Kaiming He
- Venue
- arXiv preprint — not yet peer-reviewed
- Published
- 2026-10-01
- ID
- arXiv:2610.02200
Preprint, not yet peer-reviewed. Introduces a visual harness that lets multimodal models retain and retrieve previous observations during interactive tasks. The authors report improvements on visual games and puzzles with the same underlying model; these are research benchmark results, not independent production validation.
#multimodal#visual-reasoning#ai-agents
arxiv.orgAdded Nov 7, 20260 opens 
alphaxiv.orgResearch PaperTechnology & Ethics
- Authors
- Mistral AI
- Venue
- alphaXiv (preprint, not peer-reviewed)
- Published
- 2026-11-03
- ID
- 2611.mistral-large-4
The abstract introduces Mistral Large 4 as a multimodal mixture-of-experts model for reasoning and long-horizon agentic tasks. The page presents a paper record and a lengthy list of contributors.
alphaxiv.orgAdded Nov 4, 20260 opens 
csinva.ioResearch PaperTechnology & Ethics
- Authors
- Chandan Singh
- Published
- 2026-11-02
An interactive report showing small transformer networks whose weights were written directly by an AI agent, with no training, to solve toy tasks such as bit-string parity. Every weight matrix is displayed, and code, run logs and saved programs are published on GitHub.
#weight-space-programming#interpretability#machine-learning
csinva.ioAdded Nov 3, 20260 opens 
rikiyat.github.ioResearch PaperTechnology & Ethics
- Venue
- Blog post (not peer-reviewed)
- Published
- 2026-11-02
A research blog post from MIT CSAIL researchers reporting that frontier language models can identify what a neural network was trained to do from its raw weights and can write small working networks directly. The post is not yet peer-reviewed; results are the authors' own.
#weight-space-programming#interpretability#llms
rikiyat.github.ioAdded Nov 3, 20260 opens 
journals.sagepub.comResearch PaperTechnology & Ethics
- Authors
- Anne J. Maheux, Chelly Maes
- Venue
- Perspectives on Psychological Science (peer-reviewed)
- Published
- 2026-08-28
- ID
- 10.1177/17456916261460771
An open-access review synthesizing research on how teenagers' use of AI chatbots such as ChatGPT may affect social belonging, emotion regulation, autonomy and identity, with possible benefits and risks, and a call for long-term developmental studies.
#adolescents#generative-ai#psychology
journals.sagepub.comAdded Nov 3, 20260 opens arxiv.org Research PaperTechnology & Ethics
- Authors
- Aaron Reich, Diana Wolfe, Matt Price, Alice Choe, Fergus Kidd, Hannah Wagner
- Venue
- arXiv (preprint, not yet peer-reviewed)
- Published
- 2026-02-26
- ID
- arXiv:2602.23279
Preprint examining whether psychological safety is associated with workplace AI adoption and usage, using survey data from 2,257 employees at a global consulting firm.
#ai-adoption#workplace#psychological-safety
arxiv.orgAdded Oct 29, 20260 opens 
dl.acm.orgResearch PaperTechnology & Ethics
An ACM paper examining how conversational AI memory features build what the authors call an "algorithmic self-portrait" from information users disclose in private chats, and the questions this raises about data sensitivity and user agency.
#AI memory#privacy#conversational AI
dl.acm.orgAdded Oct 22, 20260 opens 
cs.cmu.eduResearch PaperTechnology & Ethics
Ken Thompson's 1984 Turing Award lecture, which shows how a compiler can be altered to insert a hidden backdoor that survives even when the source code looks clean, and asks how far any program can be trusted.
#computer security#compilers#Trojan horse
cs.cmu.eduAdded Oct 22, 20260 opens arxiv.org Research PaperTechnology & Ethics
An arXiv paper (accepted to EMNLP 2026) introducing Cross-Psych-101, a dataset of 591 participants and 261,551 decisions across 12 tasks, and a model trained to predict a person's choices in one task from their behavior in another.
#LLMs#psychology#decision-making
arxiv.orgAdded Oct 20, 20260 opens arxiv.org Research PaperTechnology & Ethics
arXiv preprint (July 2026) proposing a causal-inference framing for choosing data mixtures in language-model training.
#language models#training data#causal inference
arxiv.orgAdded Oct 19, 20260 opens 
media.mit.eduResearch PaperTechnology & Ethics
MIT Media Lab project on AI and false memories.
In plain terms: An MIT Media Lab research project studying how conversations with AI chatbots can create false memories in people.
media.mit.eduAdded Oct 16, 20260 opens 
vetta.orgResearch PaperTechnology & Ethics
Shane Legg's 2008 doctoral thesis on universal intelligence.
In plain terms: Shane Legg's 2008 PhD thesis defines machine intelligence formally and examines what superintelligent machines could mean; Legg later co-founded DeepMind.
vetta.orgAdded Oct 16, 20260 opens arxiv.org Research PaperTechnology & Ethics
The paper that introduced the Transformer architecture.
In plain terms: The 2017 Google paper introducing the Transformer, the attention-based architecture behind today's large language models. Peer-reviewed (NeurIPS 2017).
arxiv.orgAdded Oct 16, 20260 opens 
pmc.ncbi.nlm.nih.govResearch PaperTechnology & Ethics
An open-access research article investigating the relationship between the use of generative AI tools and self-reported depressive symptoms among adults in the United States, providing empirical data on the mental health implications of AI adoption.
pmc.ncbi.nlm.nih.govAdded Oct 14, 20260 opens 