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. 2249
    Prev
    Next
    INDEX
    Explanations

    phrases indicative of interpersonal relationships and dialogue

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

    people

    np_acts-logits-general · gemini-2.0-flash

    pronouns after punctuation

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     которое
    -0.60
     яке
    -0.53
    urtstag
    -0.51
    trone
    -0.50
    YOND
    -0.49
     hivyo
    -0.49
    PONENTS
    -0.49
    いくつか
    -0.49
     thừa
    -0.48
    것은
    -0.47
    POSITIVE LOGITS
     who
    1.02
     whom
    0.99
    此人
    0.95
    他是
    0.93
    whom
    0.91
    who
    0.90
     he
    0.89
     she
    0.88
     shes
    0.87
     him
    0.84
    Activations Density 0.356%

    No Known Activations