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

    Neuronpedia

    Jacobian LensNEW
    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. 41-GEMMASCOPE-RES-16K
    11. 8548
    Prev
    Next
    INDEX
    Explanations

    expressions of regret or apologies

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.41.hook_resid_post
    Hook Layer
    41
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    sidemargin
    -0.40
    apnews
    -0.39
     utafitiHapana
    -0.38
    ñores
    -0.38
    cydow
    -0.35
    esor
    -0.35
    batim
    -0.34
    ostante
    -0.33
    mäh
    -0.33
     kılıf
    -0.32
    POSITIVE LOGITS
     offend
    0.89
     upsetting
    0.89
     intr
    0.82
     disturb
    0.79
     offending
    0.78
     disrupting
    0.76
     intru
    0.74
     inconven
    0.72
     disruption
    0.71
     upset
    0.71
    Activations Density 0.399%

    No Known Activations