© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 22-GEMMASCOPE-RES-16K
    6. 13173
    Prev
    Next
    INDEX
    Explanations

    commands or phrases related to the act of letting go or releasing

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

    Following "let" or "letting"

    np_acts-logits-general · gemini-2.0-flash

    letting go, being loose

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Estadística
    -0.63
    Personensuche
    -0.60
     للاسماء
    -0.60
    argout
    -0.56
    濫
    -0.56
    bayaran
    -0.55
    Composable
    -0.54
    agramm
    -0.54
    mycompany
    -0.53
    uyen
    -0.53
    POSITIVE LOGITS
     loose
    1.04
     slip
    0.97
     byg
    0.95
     go
    0.92
    tuce
    0.81
    loose
    0.80
    Loose
    0.74
     alone
    0.72
    slip
    0.70
    Slip
    0.69
    Activations Density 0.105%

    No Known Activations