© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 14697
    Prev
    Next
    INDEX
    Explanations

    as well

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

    This neuron strongly activates on occurrences of the function word “as,” particularly when it’s used to introduce comparisons or clauses.

    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
    考虑
    -1.14
     viso
    -1.09
     krep
    -1.05
    瞠
    -1.05
    ztés
    -1.03
     funer
    -1.02
     cioc
    -1.01
     wichtigsten
    -0.99
     comanda
    -0.99
    から
    -0.98
    POSITIVE LOGITS
     you
    1.53
     I
    1.40
    為に
    1.21
    ほうがいい
    1.10
    際は
    1.05
    つづく
    1.02
    太多的
    0.99
     preferencia
    0.98
     είχε
    0.97
    方がいい
    0.96
    Activations Density 0.045%

    No Known Activations