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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 31-GEMMASCOPE-RES-16K
    6. 15772
    Prev
    Next
    INDEX
    Explanations

    action-oriented phrases and words related to goal pursuit, progress, and forward movement. It activates strongly for words and phrases like "to pursue", "take action", "move toward", "accomplish", "perform tasks", "initiate", and similar concepts that indicate actively working towards objectives or making

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @aml7990

    phrases related to decision-making and personal agency

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

    preceding goals, actions, choices, or tasks

    np_acts-logits-general · gemini-2.0-flash

    choosing actions toward outcomes

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    msgTypes
    -0.66
    featureID
    -0.65
     AssemblyCompany
    -0.61
    WriteBarrier
    -0.55
     الرياضيه
    -0.54
    __*/
    -0.53
    fjspx
    -0.53
    tonode
    -0.53
    sidemargin
    -0.52
    sensi
    -0.51
    POSITIVE LOGITS
     situations
    0.59
     outcomes
    0.58
     actions
    0.57
     tasks
    0.57
     action
    0.56
     outcome
    0.54
     situation
    0.53
     task
    0.53
     things
    0.51
     circonstances
    0.48
    Activations Density 0.596%

    No Known Activations