Artificial intelligence can do remarkable things, but how much does it actually understand about the world? Researchers have found that surprisingly specific information can be extracted from the inner workings of large language models (LLMs) like ChatGPT. Using mathematical tools to probe a model's internal activity, they can identify where a city is located on a map, when a historical figure lived or what emotion a character feels, even detecting descriptions of pain.
🛡️
Just a quick checkWe’re checking your connection to prevent automated abuse
| # | Наименование новости | Тональность | Информативность | Дата публикации |
|---|---|---|---|---|
| 1 | LLMs show a human-like double standard when judging others' reasoning | 0 | 9.31 | 15-10-2026 |
| 2 | Top developers are shifting from chatbots to physical AI. Here’s why | 0 | 5 | 25-06-2026 |
| 3 | Scientists are seriously asking if bees and ChatGPT are conscious | 0 | 7.39 | 05-06-2026 |
| 4 | A case for AI models that understand, not just predict, the way the world works | 5 | 7 | 15-12-2025 |
| 5 | Hidden prompts can plant false memories in AI agents, researchers warn | 0 | 7 | 19-07-2026 |
| 6 | Hidden goals can undermine AI teamwork, study finds | 0 | 8.24 | 06-08-2026 |
| 7 | Comment les IA apprennent à si bien nous parler | 0 | 9.31 | 27-05-2026 |
| 8 | Китайский врач поразил математический мир, решив многолетнюю задачу с помощью ChatGPT | 0 | 9.07 | 14-08-2026 |
| 9 | Давеча Сэм Альтман сообщил, что его новый ChatGPT под названием ... | 0 | 7.04 | 09-09-2026 |