© 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. 328
    Prev
    Next
    INDEX
    Explanations

    negations or contractions, particularly the word "don't."

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

    contractions with "n't" (negations in the form of "don't," "doesn't," "wouldn't," "isn't," etc.).

    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
     
    -0.64
     (
    -0.57
    -
    -0.53
     $
    -0.52
     $\
    -0.47
    ↵↵
    -0.46
     "
    -0.46
     \
    -0.45
    .
    -0.45
     -
    -0.43
    POSITIVE LOGITS
     queſta
    1.22
     Administrativna
    1.20
    <unused14>
    1.20
    <unused79>
    1.20
    <unused3>
    1.19
    [@BOS@]
    1.19
    <unused8>
    1.19
    <unused17>
    1.19
    <unused52>
    1.19
    <pad>
    1.18
    Activations Density 0.009%

    No Known Activations