© 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. 95537
    Prev
    Next
    INDEX
    Explanations

    subsublimesubplotidsubcmdsubscriptsubsetssubmenusubchaptersubspacesublimesub1sub1subdirsub (version)

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

    The neuron strongly activates on any token containing the substring “sub,” effectively detecting occurrences of “sub” in words or identifiers.

    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
    .
    -2.69
     considérée
    -2.53
    掼
    -2.33
    鹮
    -2.31
    malige
    -2.30
     lutar
    -2.30
     houver
    -2.14
     spannende
    -2.08
    ៚
    -2.06
    ⮑
    -2.05
    POSITIVE LOGITS
    \
    3.11
    ed
    2.89
    ୈ
    2.78
    饜
    2.77
    ୌ
    2.70
    ⎆
    2.69
    *
    2.61
    鼉
    2.58
    section
    2.56
    u
    2.47
    Activations Density 0.052%

    No Known Activations