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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128269
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on words and short phrases that occur in giving step-by-step or how-to instructions (e.g. imperative verbs like “check,” “add,” “learn” together with their accompanying nouns and prepositions).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    later / afterwards

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     fréquemment
    -0.98
    OOOOO
    -0.89
     mêmes
    -0.81
     perle
    -0.81
    ACIN
    -0.80
     equival
    -0.79
     expliquer
    -0.79
     lohnt
    -0.79
     financiación
    -0.79
    ABOR
    -0.78
    POSITIVE LOGITS
     later
    1.49
     afterwards
    1.41
     will
    1.41
    之後
    1.23
     AFTER
    1.23
    以后
    1.19
     später
    1.18
    年後
    1.15
    以後
    1.14
    Later
    1.13
    Activations Density 0.025%

    No Known Activations