© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 11918
    Prev
    Next
    INDEX
    Explanations

    have/has followed by past participle

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron detects passive‐voice perfect constructions (forms of “has/have been” followed by a past participle).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    रा
    1.09
    后来
    1.09
    뵐
    1.07
    有很多
    1.03
    後來
    1.01
    اؤ
    1.01
    á
    1.01
    𝐨
    0.99
     owed
    0.99
    ्य
    0.98
    POSITIVE LOGITS
     been
    1.75
     sido
    1.65
     fått
    1.59
    been
    1.46
     vært
    1.38
     Been
    1.37
    Been
    1.34
     været
    1.31
     BEEN
    1.25
     become
    1.21
    Activations Density 0.932%

    No Known Activations