# Post Analysis: @thesupermannx on LLMs "obsessed with Japan" (2026-10-21, with the 2026-07-20 version) Back to [[Users/@thesupermannx/thesupermannx-x-profile|@thesupermannx profile]] · Paper note: [[Users/@thesupermannx/Papers/Why are all LLMs Obsessed with Japanese Culture (arXiv 2604.21751)|arXiv 2604.21751]] · Related pattern: [[Topics/Ribosome Pausing and Cotranslational Folding/Related Trends#1. "AI cracked open X" posts that re-label old science|Ribosome topic: the "AI cracked open X" pattern]] **Verdict:** The core finding is real. The July version is fairly faithful, but the October rewrite invents mechanisms (safety guidelines, quantization, 'reddit and anime wiki users') that the paper does not report, one of which contradicts the post itself. ## The post (verbatim): October version Posted **Oct 21, 2026, 8:08 AM CT**. Link: [arXiv 2604.21751v1](https://arxiv.org/pdf/2604.21751v1). > Researchers proved every LLM is secretly obsessed with Japan. > > And they finally figured out why. > > They published a paper investigating hidden cultural biases in large language models. > > they tested 31,000 questions across 24 different languages on models like gemini, chatgpt, claude, and llama. > > they stripped out all geographic hints and just asked the models to pick a place and answer. > > the finding was universal.. across a wide range of models and languages, there is a pervasive "obsession" with japanese culture. > > even when you prompt them in a language with zero ties to japan, they default to japanese norms, values, and aesthetics at a disproportionately high rate compared to its actual global demographic weight. > > the craziest part? they tracked down exactly when the ai develops this obsession.. and it is not from reading the raw internet. > > during pre-training, the models actually have a fairly balanced, global view of the world. the obsession gets injected during the safety phase. > > when tech companies fine-tune these models to be "helpful, harmless, and polite," the ai starts mapping those corporate safety guidelines directly onto japanese cultural concepts. in the pursuit of creating the perfect, inoffensive, compliant assistant, they warped how the models view the world. > > the quantization process (compressing models to run faster locally) actually amplifies this determinism. the smaller and faster your local model is, the more stereotypically it behaves, causing japanese cultural tropes to spike hard. > > we spend all this time worrying about existential ai safety, but the real alignment problem is that our most advanced reasoning engines are hyper-fixated on japanese internet culture. > > every model upgrade is reinforcing an invisible regional bias that dictates how they generate stories, facts, and code. the cultural commonsense of the internet is not universal.. > > it’s just localized to whatever the heaviest reddit and anime wiki users were obsessed with. ## The earlier post (verbatim): July version Posted **Jul 20, 2026, 4:09 AM CT**. [Link to the post](https://x.com/thesupermannx/status/2079131466496352449). > Researchers proved every major LLM is secretly obsessed with Japan. > > And they finally figured out why. > > For years, we’ve been told that AI is entirely Western-centric, that it just reflects Silicon Valley and American values. > > A landmark paper by Cardiff and Basque researchers tested 31,680 cultural prompts across 24 languages on frontier models like ChatGPT, Claude, and Gemini. > > The results shattered that assumption. > > In six out of eight frontier models, Japan was the single most frequently referenced country when asked open-ended cultural questions. > > Ask about traditional dances, festivals, or everyday practices in an open context, and the AI defaults to Japan. > > Over and over again. > > Here is the twist nobody expected. > > This bias doesn't come from raw pre-training internet data. > > The researchers tracked where the obsession forms. It emerges after pre-training, during the supervised fine-tuning and alignment phase when humans teach the AI how to behave. > > Why Japan? > > Because decades of global soft power, rich cultural export, and clean, universally admired digital archives make Japanese culture uniquely "safe" for AI safety filters to lean on. > > When labs train models to be harmless and universally pleasing, the AI defaults to the cultural equivalent of comfort food. > > It avoids controversy by talking about anime, sushi, and tradition. ## Engagement (Oct 24, 2026) | Impressions | Likes | Reposts | Quotes | Replies | Bookmarks | |---:|---:|---:|---:|---:|---:| | 2,424,854 | 10,087 | 2,274 | 797 | 305 | 6,238 | | 3,757,239 | 10,200 | 2,040 | 778 | 227 | 4,758 | The first row is October and the second is July. Together they are the account's most-viewed paper. ## Claim-by-claim check **Ratings:** ✅ accurate · 🟡 overstated or simplified · 🔀 misattributed · ❓ unsupported by the paper · ❌ contradicted by the paper | # | Claim | Rating | What the sources show | |---|---|---|---| | 1 | "Researchers proved every LLM is secretly obsessed with Japan" | 🟡 | Japan tops most models (6 of 8 in one passage; 5 Japan / 2 US / 1 tie in the table), not "every LLM" | | 2 | 31,000 / 31,680 questions, 24 languages | ✅ | CROQ dataset | | 3 | (Jul) "Cardiff and Basque researchers" | ✅ | Cardiff University and HiTZ (EHU) | | 4 | (Jul) "six out of eight frontier models" | 🟡 | The paper's own figures are inconsistent (6/8 vs 5/2/1). Also, the 8 models are not all "frontier" ⚠️ | | 5 | Not from pre-training; emerges at fine-tuning | ✅ | The bias appears after pre-training, at SFT | | 6 | (Oct) "the obsession gets injected during the safety phase"; models map "corporate safety guidelines" onto Japanese concepts | ❓ | The paper places it at supervised fine-tuning. The safety-guideline mechanism is the post's invention | | 7 | (Oct) "the quantization process … amplifies this" | ❓ ⚠️ | Not reported in the paper | | 8 | (Oct) "localized to whatever the heaviest reddit and anime wiki users were obsessed with" | ❌ | This contradicts the paper and the post's own line that it is "not from reading the raw internet" | | 9 | (Jul) "Because decades of global soft power…" | ❓ | The post's speculation, presented as the answer | ## What the post leaves out - **The size of the effect:** how large it is per model. - **The paper's own inconsistency:** the 6/8 versus 5/2/1 counts. - **The authors' explanations:** what they actually suggest as causes. ## Bottom line - **A textbook drift case:** the same paper is reposted three months later. - **What changed:** the second version swaps cautious wording for new, unsupported mechanisms, and it went viral again (2.4M impressions).