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-2-9B
    4. Residual Stream - 16k
    5. 36-GEMMASCOPE-RES-16K
    6. 11552
    Prev
    Next
    INDEX
    Explanations

    terms related to criminal activity and escape narratives

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.36.hook_resid_post
    Hook Layer
    36
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    IntoConstraints
    -0.59
     $_"
    -0.48
    expandindo
    -0.47
    XmlRootElement
    -0.46
    ghest
    -0.45
     gynhyrchwyd
    -0.44
    WebServlet
    -0.44
    󠁴
    -0.43
    RotationOrder
    -0.43
    adaptiveStyles
    -0.43
    POSITIVE LOGITS
     escape
    2.94
     escaping
    2.56
     escapes
    2.55
     escaped
    2.45
     Escape
    2.38
    escape
    2.34
    Escape
    2.25
     ESCAPE
    2.19
    逃
    2.13
     fleeing
    2.11
    Activations Density 0.590%

    No Known Activations