{"slug": "your-transformer-can-hold-two-thoughts-at-once-evidence-of-linear-superposition", "title": "Your Transformer Can Hold Two Thoughts at Once: Evidence of Linear Superposition in LLMs", "summary": "Researchers demonstrated that Large Language Models exhibit fundamental linearity, producing a superposition of individual next-token distributions when inputs from distinct text streams are linearly combined, a phenomenon they term \"S\" (as named in the work). The finding indicates that despite relying on highly non-linear components, LLMs can hold two thoughts at once through linear superposition of next-token distributions.", "body_md": "While Large Language Models (LLMs) rely on highly non-linear components, in this work we demonstrate that they exhibit fundamental linearity: when inputs from distinct text streams are linearly combined, the model outputs a superposition of the individual next-token distributions. We term this the S", "url": "https://wpnews.pro/news/your-transformer-can-hold-two-thoughts-at-once-evidence-of-linear-superposition", "canonical_source": "https://aiflash.com/news/126181/", "published_at": "2026-09-25 13:30:14+00:00", "updated_at": "2026-09-25 13:59:32.666367+00:00", "lang": "en", "topics": ["large-language-models", "ai-research", "machine-learning", "artificial-intelligence"], "entities": [], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/your-transformer-can-hold-two-thoughts-at-once-evidence-of-linear-superposition", "markdown": "https://wpnews.pro/news/your-transformer-can-hold-two-thoughts-at-once-evidence-of-linear-superposition.md", "text": "https://wpnews.pro/news/your-transformer-can-hold-two-thoughts-at-once-evidence-of-linear-superposition.txt", "jsonld": "https://wpnews.pro/news/your-transformer-can-hold-two-thoughts-at-once-evidence-of-linear-superposition.jsonld"}}