© 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. 124674
    Prev
    Next
    INDEX
    Explanations

    unique suffixes or patterns in words, particularly in a variety of languages or contexts

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

    parts of words ending in "ing" or preceded by common words

    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
     étoient
    -0.66
     chrétien
    -0.62
     avoient
    -0.60
     feroit
    -0.59
    hésite
    -0.58
     rubio
    -0.57
    skrift
    -0.55
    étoit
    -0.54
     mijne
    -0.54
     مشين
    -0.53
    POSITIVE LOGITS
    kuuta
    0.59
     में
    0.58
     in
    0.57
    ில்
    0.56
    országban
    0.55
    માં
    0.54
    esinde
    0.53
    Тут
    0.53
     yılında
    0.53
    த்தில்
    0.52
    Activations Density 0.054%

    No Known Activations