© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 11863
    Prev
    Next
    INDEX
    Explanations

    names following initials

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

    The neuron specifically detects press-release distribution tags like “(GLOBE NEWSWIRE)” or “Marketwired” that appear in newswire datelines.

    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
    érons
    0.35
    ﺴ
    0.33
     Ř
    0.32
     Católica
    0.32
    ención
    0.32
     Các
    0.32
     Mereka
    0.32
     Từ
    0.31
     秦
    0.31
     Güzel
    0.31
    POSITIVE LOGITS
     hardened
    0.27
     untested
    0.25
     lets
    0.24
     analysts
    0.24
     spared
    0.24
     let
    0.23
    ―
    0.23
     inexperienced
    0.23
     isn
    0.23
     airing
    0.23
    Activations Density 0.000%

    No Known Activations