© 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-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 106392
    Prev
    Next
    INDEX
    Explanations

    mathematical symbols and numerical values

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

    mathematical expressions

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

    LaTeX math array/matrix and aligned-equation structures, especially vector/matrix rows and alignment points within multi-line formulas.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    tagHelperRunner
    -1.42
     queſta
    -1.22
    setVerticalGroup
    -1.20
    niſſe
    -1.17
    ロウィン
    -1.16
    ſſung
    -1.14
    ſicht
    -1.13
    iſen
    -1.13
    <unused41>
    -1.12
    [@BOS@]
    -1.11
    POSITIVE LOGITS
    <td>
    0.73
    The
    0.66
    (
    0.55
    :
    0.54
     the
    0.53
    [toxicity=0]
    0.53
    hline
    0.52
    <th>
    0.49
    0
    0.49
    _
    0.49
    Activations Density 0.002%

    No Known Activations