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. 15794
    Prev
    Next
    INDEX
    Explanations

    phrases describing ideal or perfect situations or objects

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

    Preceding "for" or describing something "ideal"

    np_acts-logits-general · gemini-2.0-flash

    ideal choice or candidate

    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
     gebou
    -0.37
    Gameplay
    -0.37
    readlines
    -0.36
    vellous
    -0.35
    bru
    -0.35
     Tun
    -0.34
    genodigd
    -0.34
    terus
    -0.34
    cinogen
    -0.33
    weer
    -0.32
    POSITIVE LOGITS
     choice
    0.92
     suited
    0.79
    choice
    0.73
    suited
    0.71
     candidate
    0.69
    Choice
    0.67
     Choice
    0.67
     CHOICE
    0.66
     candidates
    0.65
     option
    0.64
    Activations Density 0.189%

    No Known Activations