© 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. 4735
    Prev
    Next
    INDEX
    Explanations

    the presence of specific numerical and categorical data in structured contexts

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

    Followed by numbers

    np_acts-logits-general · gemini-2.0-flash

    quantifiers and amounts

    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
    ########.
    -0.74
    farande
    -0.62
    AutoScaleMode
    -0.60
     himſelf
    -0.60
     Theſe
    -0.59
    ніципалі
    -0.58
    unknownFields
    -0.56
    AutoScale
    -0.56
     themſelves
    -0.55
     Inscrivez
    -0.55
    POSITIVE LOGITS
     وفي
    0.51
     arenas
    0.49
     arena
    0.47
     context
    0.44
     fields
    0.43
     in
    0.42
     regard
    0.42
     וב
    0.42
     lieu
    0.42
     areas
    0.42
    Activations Density 1.256%

    No Known Activations