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. 419
    Prev
    Next
    INDEX
    Explanations

    topics related to balancing work and family responsibilities, particularly for women. It seems to activate strongly on phrases discussing career advancement, childcare, pregnancy, work-life balance, and challenges women face in managing professional and personal obligations.

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @jacktpayne51

    Phrases describing relationships between groups or concepts, often using prepositions like "in", "to", and "with" to connect ideas or specify contexts.

    eleuther_acts_top20 · claude-3-5-sonnet-20240620Triggered by @jacktpayne51

    topics related to gender roles and women's careers in relation to family planning

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

    Work-life balance and family planning

    np_acts-logits-general · gemini-2.0-flash

    balancing work and family commitments

    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
     fij
    -0.57
    UnusedPrivate
    -0.54
    élie
    -0.50
    AndEndTag
    -0.50
    geted
    -0.48
     Branca
    -0.48
     trúc
    -0.47
     agré
    -0.46
    alda
    -0.46
    SerializedSize
    -0.46
    POSITIVE LOGITS
     juggling
    0.75
     schedules
    0.73
     multitasking
    0.72
    tvguidetime
    0.72
     excused
    0.72
     schedule
    0.69
     prioriti
    0.68
     priorities
    0.68
     postpon
    0.67
     scheduling
    0.67
    Activations Density 0.473%

    No Known Activations