© 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. 1723
    Prev
    Next
    INDEX
    Explanations

    specific terms related to criminal activities and their consequences

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

    legal and criminal proceedings involving murder, conspiracy, and detailed case descriptions.

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

    language that attributes cause, responsibility, or participation for outcomes, especially harmful events.

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

    language describing causation, culpability, or attribution of harmful events, actions, or medical conditions.

    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
     المعيارى
    -0.46
    principalTable
    -0.44
     prejud
    -0.43
     replay
    -0.43
    webElementGuid
    -0.42
     disp
    -0.41
    Bibitem
    -0.41
     hochzeit
    -0.41
     port
    -0.40
    erevan
    -0.40
    POSITIVE LOGITS
     apparently
    0.56
    apparently
    0.52
    archiviato
    0.51
     probabilmente
    0.48
     apparent
    0.46
    probably
    0.46
     probably
    0.42
     evidently
    0.42
     probablemente
    0.41
    DockStyle
    0.41
    Activations Density 0.883%

    No Known Activations