© 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. 129412
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on very common high‐frequency words (function words and generic nouns) rather than on rarer, content-specific terms.

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

    only option or best chance

    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
     cacher
    -1.07
    ætter
    -1.06
    じゃなくて
    -1.01
    Ἴ
    -0.99
     rafra
    -0.97
     asistente
    -0.96
     菇
    -0.95
     ſont
    -0.95
     esat
    -0.95
    ulement
    -0.94
    POSITIVE LOGITS
     available
    1.72
     for
    1.35
    唯一
    1.26
     доступ
    1.21
     After
    1.19
     Only
    1.19
     after
    1.13
     provide
    1.12
     accessible
    1.09
     access
    1.09
    Activations Density 0.029%

    No Known Activations