© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 26868
    Prev
    Next
    INDEX
    Explanations

    instances of negation or inability, particularly the word "can't."

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot

    cannot followed by verb

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Negative Logits
     endwhile
    -0.40
     HasFactory
    -0.39
     okuyayım
    -0.36
    ;++
    -0.35
     investigation
    -0.35
    ☆☆☆
    -0.34
     tienden
    -0.34
     mourut
    -0.34
     nakalista
    -0.33
    NewGuid
    -0.33
    POSITIVE LOGITS
     cant
    0.89
    Cant
    0.82
    Cannot
    0.82
     Cannot
    0.81
    cannot
    0.80
    t
    0.77
    CANT
    0.77
     cannot
    0.76
     Cant
    0.74
    cant
    0.73
    Activations Density 0.026%

    No Known Activations