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-2-2B
    4. Residual Stream - 16k
    5. 24-GEMMASCOPE-RES-16K
    6. 15353
    Prev
    Next
    INDEX
    Explanations

    phrases or terms related to legal and political accountability

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

    Something defies/interferes/hurts

    np_acts-logits-general · gemini-2.0-flash

    harmful acts and consequences

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.24.hook_resid_post
    Hook Layer
    24
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     @"/
    -0.61
     jsPsych
    -0.53
    szy
    -0.50
    ValueStyle
    -0.50
    IsMutable
    -0.48
     itertools
    -0.48
    pement
    -0.47
    Told
    -0.45
    foria
    -0.45
    Hozzáférés
    -0.45
    POSITIVE LOGITS
     harms
    0.87
     hurt
    0.80
     harmed
    0.78
     hurts
    0.76
     harm
    0.74
     endanger
    0.73
     humili
    0.72
    contentLoaded
    0.70
    harm
    0.70
     damages
    0.70
    Activations Density 0.516%

    No Known Activations