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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    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. 881
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "fun" and its variations

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

    fun followed by specific 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
    argli
    -0.50
     Metzger
    -0.49
    ecko
    -0.49
    zeile
    -0.49
    vertret
    -0.49
    bihi
    -0.47
    abyte
    -0.47
     Treiber
    -0.47
     visst
    -0.46
    そこ
    -0.46
    POSITIVE LOGITS
    FUN
    1.04
    Fun
    1.03
     Fun
    1.02
    fun
    1.00
     FUN
    0.96
     fun
    0.95
     functools
    0.70
     getF
    0.70
    URLException
    0.63
    funcs
    0.60
    Activations Density 0.005%

    No Known Activations