© 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-IT
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 7213
    Prev
    Next
    INDEX
    Explanations

    instances of harassment, abuse, and derogatory language towards individuals

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

    Terms related to mistreatment

    np_acts-logits-general · gemini-2.0-flash

    harassment and bullying

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    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
     otomatig
    -0.49
    isseur
    -0.40
    畢竟
    -0.40
    APOLIS
    -0.39
    RTSN
    -0.38
    ⤹
    -0.36
    quilibrium
    -0.35
    estination
    -0.35
    ۪
    -0.35
     Paesi
    -0.35
    POSITIVE LOGITS
     harassment
    0.63
     insults
    0.63
     Harassment
    0.61
     bullying
    0.61
    bullying
    0.60
     Bullying
    0.54
     bully
    0.54
     pinulongan
    0.52
     harass
    0.52
     bullied
    0.52
    Activations Density 0.284%

    No Known Activations