{"total":23,"papers":[{"id":"url:ai-2027.com","title":"AI 2027 (Kokotajlo)","date":"2025-04-03","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/3fd447e0ff58689ae00a4666e28bbd7faecd472dec7195ab1d524cb5402a1766/","source_url":"https://ai-2027.com/","text_available":false},{"id":"url:alignmentforum.org/posts/nffst5mio7bcaqhpa/an-extremely-opinionated-annotated-list-of-my-favourite-1","title":"An Extremely Opinionated Annotated List of My Favourite Mechanistic Interpretability Papers v2 (Alignment Forum)","date":"2024-07-07","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/28f1f482044fd8bff4e1190eded455473b6882f5493ef64f40f8ab78ec9ba79b/","source_url":"https://www.alignmentforum.org/posts/NfFST5Mio7BCAQHPA/an-extremely-opinionated-annotated-list-of-my-favourite-1","text_available":false},{"id":"url:transformer-circuits.pub/2025/attribution-graphs/biology.html","title":"Attribution graphs, the biology of the model","date":"2025-03-27","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/8b3e14c537e89eed9a4402e9c14094d69913ed24d45fdab1a82ba4b8280660e2/","source_url":"https://transformer-circuits.pub/2025/attribution-graphs/biology.html","text_available":false},{"id":"url:transformer-circuits.pub/2025/attribution-graphs/methods.html","title":"Circuit tracing / attribution graphs (methods)","date":"2025-03-27","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7d4ba254716fd13d3a1ffb6518788e304eaa7077bd9d7bce201094eb293518b4/","source_url":"https://transformer-circuits.pub/2025/attribution-graphs/methods.html","text_available":false},{"id":"arxiv:2308.08708","title":"Consciousness in Artificial Intelligence: Insights from the Science of Consciousness","date":"2023-08-17","section":"AI safety and consciousness","quality":-0.0297,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/e7d3e0c10e0ae2d2d2d5a65d4df9b37db6e65fe2ef82355a9235738ff15f907e/","source_url":"https://arxiv.org/abs/2308.08708","text_available":true},{"id":"arxiv:2501.18837","title":"Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming","date":"2025-01-31","section":"AI safety and consciousness","quality":0.128536,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/c73837c3edb5465ee3f7cc7c6e16d23e16bdbdef5e1448aeaad2135dfd87de07/","source_url":"https://arxiv.org/abs/2501.18837","text_available":true},{"id":"arxiv:2303.07103","title":"Could a Large Language Model be Conscious?","date":"2023-03-04","section":"AI safety and consciousness","quality":-0.249629,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/066af7a709204ed7ece7130b13781997ec7b3a9be7f1881c6fc2bd9aa00cea33/","source_url":"https://arxiv.org/abs/2303.07103","text_available":true},{"id":"arxiv:2502.03407","title":"Detecting Strategic Deception Using Linear Probes","date":"2025-02-05","section":"AI safety and consciousness","quality":-0.38798,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/8297cbfb305547d6946bba3998d8921b9578398ed371961d3f0eeadd871e4a49/","source_url":"https://arxiv.org/abs/2502.03407","text_available":true},{"id":"url:transformer-circuits.pub/2025/introspection","title":"Emergent Introspective Awareness in Large Language Models (Anthropic, October 2025)","date":"2025-10-29","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/30fa339b938a47b599e7cd074489162aa385bdb2e2d7a9a6f134d3a3a1bb5487/","source_url":"https://transformer-circuits.pub/2025/introspection/index.html","text_available":false},{"id":"arxiv:2603.19426","title":"Is Evaluation Awareness Just Format Sensitivity? Limitations of Probe-Based Evidence under Controlled Prompt Structure","date":"2026-03-19","section":"AI safety and consciousness","quality":-0.292022,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7f8dc25d398b7b7c0016c33d829b3613ec6d66554ba5b49dbfc2362ce8621032/","source_url":"https://arxiv.org/abs/2603.19426","text_available":true},{"id":"arxiv:2206.13353","title":"Is Power-Seeking AI an Existential Risk?","date":"2022-06-16","section":"AI safety and consciousness","quality":0.002809,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/21192da140d682c8aacb0e8dfccbbf481a855b35ee245ee548f905eb993754d4/","source_url":"https://arxiv.org/abs/2206.13353","text_available":true},{"id":"url:anthropic.com/news/mapping-mind-language-model","title":"Mapping the mind of a large language model (Anthropic)","date":"2024-05-21","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/edf322b5841fac48b6247e69ebcb8c3a4598b4af752c1d567c0f9392b778738d/","source_url":"https://www.anthropic.com/news/mapping-mind-language-model","text_available":false},{"id":"arxiv:1912.01683","title":"Optimal Policies Tend to Seek Power","date":"2019-12-03","section":"AI safety and consciousness","quality":0.047714,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7fe89b6251f3cacab8ed2ad8b728272f92e704949d342d0ab8b0ea97548860cb/","source_url":"https://arxiv.org/abs/1912.01683","text_available":true},{"id":"arxiv:2503.16348","title":"Palatable Conceptions of Disembodied Being","date":"2025-03-20","section":"AI safety and consciousness","quality":-0.72055,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/00f41876b2b8852e1d5e080ecee9244ad7ccd7501df38c0527112912d3b462cf/","source_url":"https://arxiv.org/abs/2503.16348","text_available":true},{"id":"arxiv:2206.13477","title":"Parametrically Retargetable Decision-Makers Tend To Seek Power","date":"2022-06-27","section":"AI safety and consciousness","quality":-0.098112,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/e47c3ca813aa251e657f4fd4b126d747e91e7518bcab8f7e9139a9cf5c8d95fb/","source_url":"https://arxiv.org/abs/2206.13477","text_available":true},{"id":"arxiv:2304.06528","title":"Power-seeking can be probable and predictive for trained agents","date":"2023-04-13","section":"AI safety and consciousness","quality":-0.33811,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/d7ff8707221493f94b2bbe97040d72c861160d33c5bfcce90eebf1e0ee3d3af8/","source_url":"https://arxiv.org/abs/2304.06528","text_available":true},{"id":"url:anthropic.com/research/reasoning-models-dont-say-think","title":"Reasoning Models Don't Always Say What They Think","date":"2025-05-08","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/4ca0654119289c784e751677f3ce63742c77f58afe54849ad74eafb2d1bccb8a/","source_url":"https://www.anthropic.com/research/reasoning-models-dont-say-think","text_available":false},{"id":"url:situational-awareness.ai","title":"Situational Awareness (Aschenbrenner)","date":"2024-05-29","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/b8e91c39d818c455f92112f6b1e49e78cc51f470d5af221143e7425ac3e74539/","source_url":"https://situational-awareness.ai","text_available":false},{"id":"arxiv:2309.08600","title":"Sparse Autoencoders Find Highly Interpretable Features in Language Models","date":"2023-09-15","section":"AI safety and consciousness","quality":0.096043,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/311cc8e986e42947a9cd36e64cb7dad93d7352e1f5ad495664fce8d0f6fca5da/","source_url":"https://arxiv.org/abs/2309.08600","text_available":true},{"id":"arxiv:2606.17056","title":"The Value Axis: Language Models Encode Whether They're on the Right Track","date":"2026-06-15","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/969b67b08341367524f91a84c873d4fedbe44a95fa24841e03b2f8e111caa9a6/","source_url":"https://arxiv.org/abs/2606.17056","text_available":true},{"id":"url:anthropic.com/research/tracing-thoughts-language-model","title":"Tracing thoughts of a language model / Circuit tracing (Anthropic)","date":"2025-03-27","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/6817b4eca1b10533ef10b19c81e7c61b3877f963641fea2a81c5a46f68a233cf/","source_url":"https://www.anthropic.com/research/tracing-thoughts-language-model","text_available":false},{"id":"url:openai.com/research/weak-to-strong-generalization","title":"Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision","date":"2023-12-14","section":"AI safety and consciousness","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/fb7b23a617dce974dbb70db824c97c1de84afa11e1e290d5cb9f5bd88857e762/","source_url":"https://openai.com/research/weak-to-strong-generalization","text_available":false},{"id":"arxiv:2607.23379","title":"When Activation Oracles Learn Not to Read: Concept-Specific Blind Spots in Fine-Tuned Oracles","date":"2026-07-25","section":"AI safety and consciousness","quality":0.085385,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/a90f5a062f6ed3f603a48433557f8d622d5e5e5007884638f465243b87f67ac5/","source_url":"https://arxiv.org/abs/2607.23379","text_available":true}],"next":null}