© 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 - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 12397
    Prev
    Next
    INDEX
    Explanations

    actions related to measurement, evaluation, and analysis in research contexts

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

    followed by verbs like solve, study, investigate.

    np_acts-logits-general · gemini-2.0-flash

    action verbs followed by specific outcomes

    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
     Walkover
    -0.66
    UserScript
    -0.62
     Monfieur
    -0.61
    Microkernel
    -0.59
    ロウィン
    -0.57
    AllAfrica
    -0.57
     ModelRenderer
    -0.56
     SwitchCompat
    -0.56
    ագրություններ
    -0.56
    — 
    -0.55
    POSITIVE LOGITS
     your
    0.35
     prevent
    0.33
     them
    0.33
    join
    0.32
    addGap
    0.32
    ToFit
    0.32
    #
    0.31
     ujednoznacz
    0.31
     its
    0.30
     pinulongan
    0.30
    Activations Density 0.406%

    No Known Activations