© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 106
    Prev
    Next
    INDEX
    Explanations

    finding out

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

    The neuron fires strongly on semantically rich content words—especially main verbs (often in –ing form) and other non‐function words—while ignoring common function words.

    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
     this
    -1.24
     if
    -1.14
     этом
    -1.10
     but
    -1.09
     этого
    -1.06
     цьому
    -0.99
    ↳
    -0.97
     This
    -0.96
    
    -0.96
    ็ด
    -0.94
    POSITIVE LOGITS
     a
    1.80
     also
    1.21
    んでる
    1.14
     basically
    1.13
     akin
    1.09
     something
    1.04
     yourself
    1.04
     an
    1.04
     myös
    1.04
     roughly
    1.02
    Activations Density 0.023%

    No Known Activations