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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 75238
    Prev
    Next
    INDEX
    Explanations

    the term "defendant" in various contexts

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

    defendant's legal context

    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
     kasarigan
    -0.78
    seamnă
    -0.61
     ältere
    -0.58
    Malley
    -0.57
    httphttps
    -0.57
    MÁS
    -0.56
     Comprometido
    -0.56
     älteren
    -0.53
     scoop
    -0.53
     scoops
    -0.52
    POSITIVE LOGITS
     defendant
    1.07
     Defendant
    1.03
    Defendant
    0.90
    defendant
    0.89
     DEFENDANT
    0.78
     defendants
    0.73
     Defendants
    0.69
    StoryboardSegue
    0.52
     respondent
    0.50
     contestant
    0.50
    Activations Density 0.042%

    No Known Activations