Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 95276
    Prev
    Next
    INDEX
    Explanations

    keywords and phrases indicating the presence of magical or mystical concepts and events

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

    divine revelation, increase variance, partial effects, cell transporter, tunnel current, air systems, microcomputer, somatic symptoms

    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
     Administrativna
    -0.60
     feroit
    -0.56
    ImageContext
    -0.53
     étoient
    -0.51
     saites
    -0.48
     pouvoit
    -0.48
     étoit
    -0.47
    oa̍t
    -0.46
    providedIn
    -0.46
     Numerade
    -0.45
    POSITIVE LOGITS
    ⓧ
    0.51
    évaluateur
    0.44
    prechpartner
    0.43
    HtmlAttribute
    0.43
     nantinya
    0.41
    centralwidget
    0.40
    FunctionFlags
    0.38
     Ternyata
    0.36
    awarkan
    0.36
     wiederum
    0.35
    Activations Density 0.159%

    No Known Activations