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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 15472
    Prev
    Next
    INDEX
    Explanations

    scientific measurements and analyses related to experimental processes

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

    Past tense verbs in passive voice

    np_acts-logits-general · gemini-2.0-flash

    actions completed

    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
    Према
    -0.76
    MLLoader
    -0.74
     Efq
    -0.73
     fubject
    -0.71
     Diſ
    -0.71
    TagMode
    -0.70
    XmlAccessType
    -0.70
    vician
    -0.68
     Reſ
    -0.67
    AsUp
    -0.67
    POSITIVE LOGITS
     done
    2.16
     performed
    2.00
    done
    1.80
     conducted
    1.69
     Done
    1.67
    Done
    1.65
     Performed
    1.61
     undertaken
    1.56
    performed
    1.52
     DONE
    1.47
    Activations Density 0.788%

    No Known Activations