© 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
    7. Residual Stream - 16k
    9. 31-GEMMASCOPE-RES-16K
    11. 12618
    Prev
    Next
    INDEX
    Explanations

    elements of analysis and evaluation in scientific research and discussions

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

    connections between ideas using the word "and" followed by a description or explanation of consequences.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    words that connect clauses or list items with "and" in academic or formal writing contexts.

    oai_token-act-pair · claude-4-5-haikuTriggered by @emiglarou
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_31/width_16k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     réalis
    -0.52
    WindowConstants
    -0.51
    Personensuche
    -0.51
     considér
    -0.47
     apó
    -0.45
    +:+
    -0.40
    accumulative
    -0.39
     élève
    -0.38
     ſche
    -0.38
    RefNanny
    -0.37
    POSITIVE LOGITS
     taken
    1.26
     analyzed
    1.14
     examined
    1.13
     evaluated
    1.13
     thrown
    1.12
     brought
    1.12
     transferred
    1.12
     inspected
    1.12
     treated
    1.11
     eaten
    1.09
    Activations Density 0.910%

    No Known Activations