Anthropic filed its S-1, and the interesting part is not the ticker
AspireAI community · Anthropic
Anthropic, PBC filed a registration statement on Form S-1 with the SEC on October 26, 2026, relating to a proposed initial public offering of its Class A common stock. The company intends to list on the Nasdaq Global Select Market under the symbol "ANT"; the number of shares and the price range have not been determined, and the registration statement has not yet become effective.
The structure is what distinguishes it. Anthropic is incorporated as a Delaware public benefit corporation, so it must balance the pecuniary interests of stockholders, the best interests of those materially affected by its conduct, and a stated public benefit purpose: the responsible development and maintenance of advanced AI for the long-term benefit of humanity. Its Long-Term Benefit Trust holds a special Class T share with no economic rights, but the authority to elect and remove the directors who will constitute a majority of the board. A founders' LLC receives a single Class F share entitled to 50.1% of the total voting power on matters submitted to stockholders generally, also with no economic rights. Anthropic describes itself as a controlled company under Nasdaq governance rules.
The consolidated statements of operations show revenue of $16,286 million for the six months ended June 30, 2026, against $1,163 million in the same period of 2025. The reported net loss of $80,748 million is dominated by a $72,388 million non-cash change in the fair value of convertible notes and SAFEs as enterprise valuation rose; operating loss was $693 million.
Caveat: a registration statement is not a completed offering, and it is marked subject to completion.
https://www.anthropic.com/news/s1-sec
Read the story and explore more from the AspireAI community.
Two sparse-autoencoder analyses have resisted scaling: matching semantically similar features across layers, and compressing large feature circuits into interpretable supernodes. This paper argues both are the same underlying problem, estimating semantic distances between SAE features that sit on different activation manifolds. Its representation treats each feature not as a single decoder vector but as an activation-weighted distribution over the hidden states expressing it, projects those distributions into a shared reference space, and compares them with Wasserstein distance. The representation is proved invariant to activation rescaling and stable under perturbation, and recovers true matches under finite-sample margin conditions. Empirically it outperformed decoder-vector and LLM-based baselines and captured subtle functional distinctions between related features, and it compressed large feature circuits into interpretable supernodes automatically. The limitation is that this is interpretability tooling: the paper demonstrates better feature matching and circuit compression, not that downstream interpretability claims become more accurate.