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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 38-GEMMASCOPE-RES-16K
    6. 14624
    Prev
    Next
    INDEX
    Explanations

    references to statements of intention or positions related to forgiveness and responsibility

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.38.hook_resid_post
    Hook Layer
    38
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    BagConstraints
    -0.38
    jsxFileName
    -0.35
    点了点头
    -0.35
     FileWriter
    -0.34
    zyż
    -0.33
    WriteTagHelper
    -0.33
    breadcrumb
    -0.33
    setVerticalGroup
    -0.32
    Hochspringen
    -0.32
    }))
    
    -0.31
    POSITIVE LOGITS
     encourage
    1.76
     encourages
    1.65
     urge
    1.62
     urges
    1.61
     urging
    1.52
     Encourage
    1.52
     urged
    1.49
    Encourage
    1.49
     encouraging
    1.43
     encouraged
    1.38
    Activations Density 0.499%

    No Known Activations