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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 15156
    Prev
    Next
    INDEX
    Explanations

    themes related to power structures and authority within governance

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

    governance or control

    np_acts-logits-general · gemini-2.0-flash

    rule and authority

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    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
    stdafx
    -0.49
    abung
    -0.39
    bladet
    -0.38
    queryInterface
    -0.38
     Privacidad
    -0.37
     SALUD
    -0.36
    complexity
    -0.36
    ennifer
    -0.36
    Kedves
    -0.36
    ngOnDestroy
    -0.35
    POSITIVE LOGITS
     control
    0.73
     controlling
    0.71
     authority
    0.65
     governance
    0.64
    controlling
    0.63
     influence
    0.62
     rule
    0.60
     contrôle
    0.60
     power
    0.59
     controle
    0.59
    Activations Density 0.460%

    No Known Activations