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

    Neuronpedia

    APICircuit 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. 16-GEMMASCOPE-RES-16K
    6. 15681
    Prev
    Next
    INDEX
    Explanations

    references to specific films and character roles

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

    Detective and spy stories

    np_acts-logits-general · gemini-2.0-flash

    fictional detectives and agents

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 16-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     croissance
    -0.56
     llorando
    -0.56
     giapp
    -0.55
     Growth
    -0.55
    Growth
    -0.55
     Wachstum
    -0.54
     sobs
    -0.53
     malheure
    -0.53
    cheiden
    -0.52
     growth
    -0.52
    POSITIVE LOGITS
     detective
    1.03
    detective
    0.92
     detectives
    0.91
     crime
    0.90
     Detective
    0.86
     investigative
    0.85
     Crime
    0.83
     Interpol
    0.82
     forensic
    0.80
    🕵
    0.80
    Activations Density 0.300%

    No Known Activations