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. 91620
    Prev
    Next
    INDEX
    Explanations

    references to the character Superman and related superhero themes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
     informée
    -0.63
     surla
    -0.58
     الحره
    -0.56
     nakalista
    -0.55
    usercontent
    -0.53
    ennale
    -0.47
    총
    -0.47
    parsedMessage
    -0.47
     AspNetCore
    -0.47
     Spaß
    -0.46
    POSITIVE LOGITS
     Superman
    1.03
    Superman
    0.96
     superman
    0.80
    superman
    0.77
     Krypton
    0.46
     ſch
    0.44
    testng
    0.44
     ſte
    0.42
     cæ
    0.41
     juſ
    0.40
    Activations Density 0.005%

    No Known Activations