© 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-2-2B
    5. 22-RES-MATRYOSHKA-DC
    7. 1593
    Prev
    Next
    INDEX
    Explanations

    the word "when," especially at the start of quoted speech

    oai_token-act-pair · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    chanind/gemma-2-2b-batch-topk-matryoshka-saes-w-32k-l0-40/standard/blocks.22.hook_resid_post
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    chanind/pile-uncopyrighted-gemma-1024-abbrv-1B
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    NameInMap
    -0.65
     indeed
    -0.59
     protoimpl
    -0.59
     fact
    -0.57
     while
    -0.55
    hObject
    -0.53
     inderdaad
    -0.52
     TextAppearance
    -0.51
     EconPapers
    -0.51
     anything
    -0.50
    POSITIVE LOGITS
     dealing
    0.91
     compared
    0.88
     confronted
    0.86
    soever
    0.81
     discussing
    0.77
     considering
    0.77
     faced
    0.76
     comparing
    0.76
    compared
    0.74
    considering
    0.74
    Activations Density 0.262%

    No Known Activations