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-2-2B
    4. Residual Stream - 16k
    5. 22-GEMMASCOPE-RES-16K
    6. 3346
    Prev
    Next
    INDEX
    Explanations

    references to death and dying

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

    preceding or related to death

    np_acts-logits-general · gemini-2.0-flash

    tragic and preventable deaths

    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
     للاسماء
    -0.98
    IVEREF
    -0.76
     AssemblyCulture
    -0.72
    WithMany
    -0.67
     utafitiHapana
    -0.66
    الإنجليزية
    -0.66
    DebuggerNonUser
    -0.66
    رشف
    -0.65
    makeConstraints
    -0.63
     ligiloj
    -0.62
    POSITIVE LOGITS
     tragically
    0.77
     mysteriously
    0.66
     defending
    0.64
     trying
    0.63
     fighting
    0.63
    peace
    0.61
     prematurely
    0.61
     due
    0.61
     preventable
    0.60
     needlessly
    0.58
    Activations Density 0.189%

    No Known Activations