© 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
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 3062
    Prev
    Next
    INDEX
    Explanations

    references to television and film adaptations, particularly those with notable characters or stylistic elements

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

    proper names and descriptive comparisons between pop culture references.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi

    stylistic comparisons and name-drops that reference famous people, brands, or works (e.g., “in the style/spirit of,” “X meets Y,” quoted comparisons).

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    comparisons or analogies between two things, typically introduced by phrases like "in the spirit of," "meets," "version of," or "analogue of."

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    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
     onCancelled
    -0.42
     tesis
    -0.36
    yaç
    -0.35
    ámide
    -0.35
    новниш
    -0.34
     rubia
    -0.34
     Wesen
    -0.33
    apprendre
    -0.33
    améliorer
    -0.33
     bezpiecze
    -0.33
    POSITIVE LOGITS
     Италијани
    0.60
    ंदीखरीदारी
    0.57
     OFDb
    0.55
    AndEndTag
    0.52
    exitRule
    0.52
     nahilalakip
    0.50
     autorytatywna
    0.49
    NameInMap
    0.49
    ResponseType
    0.49
    Personensuche
    0.49
    Activations Density 0.273%

    No Known Activations