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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 1890
    Prev
    Next
    INDEX
    Explanations

    instances of strong emotional responses or significant events

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

    place names and geographic locations.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi

    capitalized sentence-initial determiners (e.g., the start of a sentence beginning with a determiner).

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    line breaks or paragraph boundaries in text.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    Prompts (Dashboard)
    24,576 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
    <pad>
    -1.00
    <unused47>
    -0.99
    <unused8>
    -0.99
    <unused79>
    -0.99
    <unused41>
    -0.99
    <unused52>
    -0.99
    <unused16>
    -0.98
    <unused23>
    -0.98
    <unused3>
    -0.98
    [@BOS@]
    -0.98
    POSITIVE LOGITS
    ↵↵
    0.43
    The
    0.38
    Though
    0.35
     Though
    0.33
    Besides
    0.32
    1
    0.32
    Furthermore
    0.30
    </i>
    0.29
    However
    0.29
    (
    0.29
    Activations Density 0.457%

    No Known Activations