Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 18-GEMMASCOPE-RES-16K
    6. 1091
    Prev
    Next
    INDEX
    Explanations

    phrases and terms related to marketing and promotion in different contexts

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

    after "of," "to," "and," or "in."

    np_acts-logits-general · gemini-2.0-flash

    old english words

    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
    tuigen
    -0.52
    वाह
    -0.47
    auru
    -0.47
    ghurt
    -0.47
     Theolog
    -0.47
    xsi
    -0.46
    ूह
    -0.46
    椀
    -0.45
    大戦
    -0.45
     geometry
    -0.45
    POSITIVE LOGITS
    ſelf
    0.88
     purpoſe
    0.77
     houſe
    0.76
     ſtate
    0.74
     pleaſure
    0.72
     greateſt
    0.71
     faſt
    0.70
    Portale
    0.70
     leaſt
    0.70
     ſmall
    0.69
    Activations Density 0.924%

    No Known Activations