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. 20-GEMMASCOPE-RES-16K
    6. 14092
    Prev
    Next
    INDEX
    Explanations

    sentences emphasizing personal connections and shared experiences.

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

    words related to anxiety and mental health.

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    phrases related to providing or presenting information, particularly in written or spoken form. It activates strongly for phrases like "brought to you", "of interest", "applied to", "given you", "ahead to", and "looking for" which all relate to conveying or offering information or ideas to

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     the
    -0.59
     indisponible
    -0.53
     Mr
    -0.53
     Sun
    -0.52
    Sun
    -0.50
     оригіналу
    -0.50
    inei
    -0.48
    ftc
    -0.48
    Mr
    -0.47
     sun
    -0.47
    POSITIVE LOGITS
     μαζί
    0.91
     conmigo
    0.88
     comigo
    0.84
     kepad
    0.77
     contigo
    0.74
     conosco
    0.72
    sidemargin
    0.70
    styleType
    0.67
     עלי
    0.66
     tome
    0.62
    Activations Density 0.313%

    No Known Activations