© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 1947
    Prev
    Next
    INDEX
    Explanations

    vanishing and collapsing

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

    This neuron responds to words describing things disappearing, vanishing, or otherwise ceasing to exist.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Berdasarkan
    -0.93
    verley
    -0.91
     ഇ
    -0.88
    rogène
    -0.88
     separación
    -0.85
    rk
    -0.84
    RTCK
    -0.84
    まだまだ
    -0.84
     directement
    -0.84
     vš
    -0.83
    POSITIVE LOGITS
     leaving
    1.99
    leaving
    1.55
     replaced
    1.55
     into
    1.45
     Leaving
    1.43
     along
    1.30
    replaced
    1.27
    Replaced
    1.26
    Leaving
    1.26
     wraz
    1.21
    Activations Density 0.065%

    No Known Activations