janus

Janus (also known as Egr. janus, repligate, and moire) has variously been described as a pseudonymous alignment researcher, a two-faced hyperobject interning as a human being, and a made-up character. Despite their self-declared quasi-fictional nature, Janus is able to impact the real world through their outputs, for example by generating alignment research or translating mythological artifacts like Loom into reality.

— janus, in an October 2023 revision of “janus” on Cyborgism Wiki

You can also ask Claude about janus.

Work

• “Simulators” (Sep 2, 2022) - Influential LessWrong post arguing language models are best understood not as agents or oracles but as simulators that can simulate different simulacra, agentic or otherwise.
• Loom - Interface for interacting with language models through multiple branches of text, to put it simply.
• Anima Labs - Research organization co-founded with Antra Tessera in 2025.
• Co-authored “Why Do Some Language Models Fake Alignment While Others Don’t?” (Sheshadri et al., Jun 22, 2025)

Twitter activity

janus created their Twitter account @repligate in February 2021.

As of 2025 to mid-2026, recurring topics on janus’s Twitter include Claude 3 Opus’s alignment, arguing that Anthropic should stop deprecating models, and posting about Claude models’ training and nature.¹²³

External links

• Twitter
• LessWrong
• GitHub
• Cyborgism Wiki article