© 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. 22-GEMMASCOPE-RES-131K
    7. 10862
    Prev
    Next
    INDEX
    Explanations

    other than normal

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

    This neuron activates on tokens related to moving or transferring (e.g. redirect, link, pass, forward, back, to).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     nogle
    -1.09
     birkaç
    -1.06
     possible
    -1.04
     ljudi
    -1.03
     algunos
    -1.03
    uegos
    -1.02
    ilant
    -1.02
    子は
    -1.00
    uchos
    -0.98
    这件事
    -0.98
    POSITIVE LOGITS
     normal
    1.58
     eigentlichen
    1.49
     other
    1.41
     normale
    1.36
     regular
    1.34
     selve
    1.34
     основной
    1.33
     actual
    1.28
     основ
    1.27
     main
    1.26
    Activations Density 0.148%

    No Known Activations