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. 22-GEMMASCOPE-RES-16K
    6. 12717
    Prev
    Next
    INDEX
    Explanations

    words related to violence and crime incidents

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

    preceding prepositions or conjunctions

    np_acts-logits-general · gemini-2.0-flash

    arrested, allegedly, woman, cell

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ftagPool
    -0.48
     cancers
    -0.47
     budu
    -0.46
    cils
    -0.45
     perspective
    -0.44
    mapTo
    -0.44
    poken
    -0.43
    illeurs
    -0.43
    agot
    -0.43
    rition
    -0.42
    POSITIVE LOGITS
    
    1.00
     AssemblyCulture
    0.83
     allegedly
    0.76
    alleg
    0.71
    RenderAtEndOf
    0.71
     StatelessWidget
    0.71
    SourceChecksum
    0.69
    httphttps
    0.69
    rungsseite
    0.69
    GeneratedValue
    0.67
    Activations Density 0.150%

    No Known Activations