© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 1320
    Prev
    Next
    INDEX
    Explanations

    phrases indicating comparisons or contrasts

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

    Prepositions "to", "from", or "of"

    np_acts-logits-general · gemini-2.0-flash

    instead of, rather than, from

    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
     instead
    -0.37
     zamiast
    -0.36
    ConstraintMaker
    -0.35
    instead
    -0.34
    раздо
    -0.33
     vainqueur
    -0.32
    rather
    -0.32
    打
    -0.31
     rather
    -0.31
     istället
    -0.31
    POSITIVE LOGITS
    rungsseite
    0.68
    
    0.59
    enumii
    0.56
     pinulongan
    0.55
    IsContent
    0.55
     يتيمه
    0.54
    findpost
    0.54
    AndEndTag
    0.52
    outSlope
    0.51
    labus
    0.50
    Activations Density 0.589%

    No Known Activations