© 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-2-2B
    5. 22-GEMMASCOPE-TRANSCODER-16K
    7. 9804
    Prev
    Next
    INDEX
    Explanations

    phrases using "how" to express a method of doing something

    oai_token-act-pair · gemini-2.0-flash

    how

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     how
    -1.18
    How
    -1.00
     How
    -0.99
    how
    -0.97
     HOW
    -0.83
     bagaimana
    -0.81
    HOW
    -0.79
     why
    -0.75
     कित
    -0.72
     combien
    -0.71
    POSITIVE LOGITS
    ]")]
    0.69
    WireFormatLite
    0.69
     they
    0.66
     efficiently
    0.65
    aarrggbb
    0.63
    มาะ
    0.63
    Diwedd
    0.63
    StoryboardSegue
    0.60
    NavController
    0.59
     doInBackground
    0.59
    Activations Density 6.378%

    No Known Activations