© 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. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 30-GEMMASCOPE-RES-16K
    11. 12478
    Prev
    Next
    INDEX
    Explanations

    instances of the word "didn't" and variations, particularly in questions and negatives

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    "did" and its capitalization variations

    np_acts-logits-general · gemini-2.0-flash

    past tense questions and statements

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 30-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_30/width_16k/average_l0_120
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.30.hook_resid_post
    Hook Layer
    30
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    rungsseite
    -0.72
     Houſe
    -0.66
     houſe
    -0.63
     pleaſure
    -0.60
    featureID
    -0.60
     estekak
    -0.60
     itſelf
    -0.59
    fillType
    -0.59
    protoimpl
    -0.57
    ientôt
    -0.57
    POSITIVE LOGITS
     Did
    0.65
    Did
    0.59
    did
    0.57
     DID
    0.56
     did
    0.55
    DID
    0.51
     originally
    0.50
     previously
    0.47
     recently
    0.44
     volna
    0.41
    Activations Density 0.096%

    No Known Activations