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. 18-GEMMASCOPE-RES-16K
    6. 14410
    Prev
    Next
    INDEX
    Explanations

    expressions related to careful consideration and analysis

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

    careful thought or attention

    np_acts-logits-general · gemini-2.0-flash

    carefully examining

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    DispatchToProps
    -0.56
    iseks
    -0.52
    icose
    -0.50
    routeProvider
    -0.49
    abcdef
    -0.48
    Portail
    -0.47
    ageo
    -0.47
    ToProps
    -0.47
    eltzer
    -0.46
    andReturn
    -0.46
    POSITIVE LOGITS
     carefully
    1.62
     Carefully
    1.32
     careful
    1.24
    caref
    1.24
     cuidadosamente
    1.19
     soigneusement
    1.12
     Careful
    1.09
     thoroughly
    1.06
     attentively
    1.05
     sorgfältig
    1.05
    Activations Density 0.169%

    No Known Activations