© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 31-GEMMASCOPE-RES-131K
    11. 38538
    Prev
    Next
    INDEX
    Explanations

    references to military ranks and titles, particularly that of "Sergeant."

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

    military rank Sergeant

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     informée
    -0.57
    存于互联网档案馆
    -0.47
     незавершена
    -0.44
    FromJson
    -0.44
    argout
    -0.43
    RTEX
    -0.43
     بيها
    -0.42
    èdia
    -0.41
    Portail
    -0.40
    ew
    -0.40
    POSITIVE LOGITS
     Sergeant
    0.73
     sergeant
    0.71
    Sgt
    0.66
    Sergeant
    0.64
     Sgt
    0.64
    guiente
    0.53
     pimienta
    0.51
     Sargent
    0.50
     serge
    0.49
    major
    0.47
    Activations Density 0.003%

    No Known Activations