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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 77342
    Prev
    Next
    INDEX
    Explanations

    instances of deception or pretense in character behavior

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

    feigned not knowing, acting coy, maintaining appearances

    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
    Not in Any Lists

    No Comments

    Negative Logits
    DoubleQuotes
    -0.45
    lapsingToolbar
    -0.44
     Einfluß
    -0.43
     Verantwort
    -0.43
     verantwoorde
    -0.38
     ajedrez
    -0.38
     laikā
    -0.37
    zugehen
    -0.37
     mempel
    -0.36
    routeProvider
    -0.36
    POSITIVE LOGITS
     pretended
    0.62
     feign
    0.57
     pretending
    0.55
     faking
    0.55
     pretense
    0.55
     faked
    0.54
    afficheront
    0.53
     pretend
    0.52
     lenker
    0.52
    faux
    0.52
    Activations Density 0.025%

    No Known Activations