© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 31-GEMMASCOPE-RES-16K
    6. 4913
    Prev
    Next
    INDEX
    Explanations

    references to aviation and piloting experiences

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

    comparisons between predicted or hypothetical values and actual or real values.

    oai_token-act-pair · claude-4-5-haikuTriggered by @emiglarou

    comparisons between something and the real or actual version of it.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    mentions of “the real” thing—i.e., contrasts between simulations/depictions and actual reality, often framing real vs. fake or model vs. real-world counterparts.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_31/width_16k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     surla
    -0.63
    expandindo
    -0.61
    sizeCache
    -0.57
    IntoConstraints
    -0.55
     intptr
    -0.54
    ="@+
    -0.53
    SharedCtor
    -0.53
     BrowserModule
    -0.53
    providedIn
    -0.49
    ViewImports
    -0.49
    POSITIVE LOGITS
     actual
    2.81
    actual
    2.52
     ACTUAL
    2.42
    Actual
    2.34
     real
    2.33
     Actual
    2.25
    ACTUAL
    2.17
     actually
    2.16
     tatsäch
    2.14
     genuine
    2.03
    Activations Density 1.806%

    No Known Activations