© 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 - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 39571
    Prev
    Next
    INDEX
    Explanations

    references to text elements and their manipulation in code

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

    method selection or replacement

    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-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    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
    AutoScaleMode
    -0.43
    tonsoft
    -0.40
    ultados
    -0.40
    клопе
    -0.36
     feminina
    -0.35
     autorizado
    -0.35
     gekomen
    -0.35
    viewWillAppear
    -0.35
     experimentado
    -0.35
    unauthorized
    -0.34
    POSITIVE LOGITS
    省市镇
    0.45
     betweenstory
    0.44
    uxxxx
    0.43
    restre
    0.42
     sts
    0.42
     pinulongan
    0.42
    évaluateur
    0.41
    ########.
    0.41
    SizeF
    0.41
    лян
    0.41
    Activations Density 0.100%

    No Known Activations