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 - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 2340
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the construction or accumulation of something

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

    building up or accumulating

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     ActionListener
    -0.41
     identically
    -0.41
     expressly
    -0.41
    ніципа
    -0.39
    LookAnd
    -0.39
     categorically
    -0.38
    foreignKey
    -0.38
    Попис
    -0.38
    optString
    -0.38
    NamedQuery
    -0.37
    POSITIVE LOGITS
     buildup
    0.92
    accum
    0.81
     accumulation
    0.73
    Accum
    0.73
     accum
    0.71
    accumulation
    0.70
     accumulated
    0.68
     Accum
    0.68
     Accumulation
    0.67
     Build
    0.66
    Activations Density 0.010%

    No Known Activations