© 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. 10-GEMMASCOPE-RES-131K
    4. 110727
    Prev
    Next
    INDEX
    Explanations

    based on

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

    The neuron consistently fires on the “based on” construction, i.e. it detects occurrences of the phrase “based on.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     čo
    -1.63
    
    -1.42
    ⢓
    -1.41
    </td>
    -1.39
    тех
    -1.30
    nehå
    -1.30
    esistenza
    -1.28
    ’
    -1.25
    AKA
    -1.24
     băr
    -1.24
    POSITIVE LOGITS
    based
    1.91
     based
    1.59
    BASED
    1.46
    Based
    1.44
    にほんブログ村
    1.41
     vlastnosti
    1.37
    Overs
    1.34
     {},
    
    1.33
    from
    1.32
    ETHING
    1.32
    Activations Density 0.099%

    No Known Activations