Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 15-GEMMASCOPE-RES-16K
    6. 4706
    Prev
    Next
    INDEX
    Explanations

    references to emotional or physical comfort

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 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
     linkovi
    -0.61
    PyErr
    -0.53
     Patrimonio
    -0.53
    Equal
    -0.51
     Exactos
    -0.50
     concurrence
    -0.49
    zise
    -0.48
    GeneratedCode
    -0.48
    évaluateur
    -0.47
    ⓧ
    -0.47
    POSITIVE LOGITS
     refuge
    0.93
     comforted
    0.92
     protection
    0.85
     comfort
    0.83
     protective
    0.82
     comforting
    0.82
     solace
    0.81
     comforts
    0.81
     shelter
    0.81
    protective
    0.80
    Activations Density 0.229%

    No Known Activations