© 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-2-9B
    4. Residual Stream - 16k
    5. 15-GEMMASCOPE-RES-16K
    6. 2227
    Prev
    Next
    INDEX
    Explanations

    phrases and expressions related to caregiving and responsibility towards others or pets

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 15-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_15/width_16k/average_l0_131
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ########.
    -0.57
     Google
    -0.41
     split
    -0.41
     double
    -0.41
    putString
    -0.40
     sangu
    -0.39
     block
    -0.39
     sign
    -0.39
    வெ
    -0.39
     decisive
    -0.38
    POSITIVE LOGITS
     care
    0.94
    Caring
    0.92
     cuidar
    0.91
     cuidados
    0.91
     Care
    0.90
     cuida
    0.88
    care
    0.84
    Care
    0.83
     cuidado
    0.82
     cared
    0.82
    Activations Density 0.198%

    No Known Activations