© 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. 12992
    Prev
    Next
    INDEX
    Explanations

    bootcamp

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

    The neuron isn’t picking out any particular content word—instead it’s acting as a positional signal, firing strongly on tokens that occur late in the sequence (i.e. at high absolute positions in the document).

    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
    
    0.87
     partenaire
    0.80
     alrededores
    0.79
    『
    0.79
    중
    0.79
    Լ
    0.77
    エリア
    0.76
     boîte
    0.76
    R
    0.75
    連携
    0.74
    POSITIVE LOGITS
    storm
    0.73
    storms
    0.70
    обе
    0.68
    াইয়
    0.66
     pro
    0.66
     hitters
    0.65
    PROP
    0.65
    specialchars
    0.64
    oused
    0.64
     selfless
    0.63
    Activations Density 0.001%

    No Known Activations