© 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-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 4457
    Prev
    Next
    INDEX
    Explanations

    phrases and terms related to common operational procedures and attributes in various contexts

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

    commonality or frequency

    np_acts-logits-general · gemini-2.0-flash

    usually and typically

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    XmlType
    -0.36
     tes
    -0.36
    HKEY
    -0.35
     potential
    -0.35
    还有
    -0.34
    potentially
    -0.34
    FAILED
    -0.33
    Failed
    -0.33
     poten
    -0.33
    пло
    -0.32
    POSITIVE LOGITS
    Usually
    1.04
     usually
    1.02
     Usually
    1.01
    Typically
    1.01
     typically
    0.98
     Typically
    0.96
     suele
    0.93
    typically
    0.92
    usually
    0.92
     suelen
    0.88
    Activations Density 0.453%

    No Known Activations