© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 38-GEMMASCOPE-RES-16K
    11. 9966
    Prev
    Next
    INDEX
    Explanations

    specific terms related to areas, aspects, or modes of action in scientific contexts

    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
     aspects
    -0.53
     notions
    -0.49
    aspects
    -0.48
     forms
    -0.47
     période
    -0.46
     purposes
    -0.44
     levels
    -0.43
     tijden
    -0.42
     periods
    -0.42
     nothing
    -0.41
    POSITIVE LOGITS
     enquiry
    0.70
     attack
    0.69
     inquiry
    0.68
     operation
    0.65
     investigation
    0.65
    原始内容存档于
    0.64
     concern
    0.63
     interaction
    0.63
     infestation
    0.61
     engagement
    0.59
    Activations Density 0.758%

    No Known Activations