© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 28-GEMMASCOPE-RES-16K
    6. 7652
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a purpose or reason

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

    Following the word "for"

    np_acts-logits-general · gemini-2.0-flash

    for reasons or examples

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.28.hook_resid_post
    Hook Layer
    28
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    開催
    -0.69
    見つ
    -0.52
    行
    -0.52
    あ
    -0.50
    Melanie
    -0.48
    大満足
    -0.47
    購入
    -0.47
    ed
    -0.46
    YN
    -0.45
    LikeLike
    -0.45
    POSITIVE LOGITS
     purposes
    0.71
     sake
    0.69
     reasons
    0.66
     example
    0.60
    reasons
    0.60
     gewöhnlich
    0.58
     instance
    0.58
    purposes
    0.57
     the
    0.55
    erun
    0.55
    Activations Density 0.570%

    No Known Activations