© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 103307
    Prev
    Next
    INDEX
    Explanations

    references to visual depth or imaging techniques related to layers and interactions

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

    Word pairs or phrases that show a relationship between elements, often involving prepositions ("of", "for", "to", "with"), conjunctions, or comparative constructions that establish relationships such as spatial positioning, contrast of properties, conditions, or capacity specifications within technical or academic text.

    eleuther_acts_top20 · claude-3-7-sonnet-20250219Triggered by @lu-christina
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    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
    LookAnd
    -0.92
     مرئيه
    -0.81
    WebElementEntity
    -0.76
     ſind
    -0.76
    RegressionTest
    -0.71
     imagui
    -0.70
    utilisons
    -0.70
    bootstrapcdn
    -0.70
     fashiola
    -0.69
     autorytatywna
    -0.68
    POSITIVE LOGITS
     fær
    0.31
     somehow
    0.28
    ęt
    0.28
    hilangan
    0.26
    amoan
    0.25
    dot
    0.25
     başlad
    0.24
     competencia
    0.24
     ausgeschlossen
    0.24
    勇
    0.24
    Activations Density 0.175%

    No Known Activations