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. 15489
    Prev
    Next
    INDEX
    Explanations

    instances of repetitive conjunctions and phrases indicating continued action or states

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

    after conjunctions "and" or "but"

    np_acts-logits-general · gemini-2.0-flash

    coordinating conjunctions

    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
     OMITBAD
    -0.43
     препратки
    -0.40
    檚
    -0.40
     themselves
    -0.39
     itſelf
    -0.38
    FieldOffsetTable
    -0.36
    themselves
    -0.36
    VolleyError
    -0.33
    Rüyada
    -0.32
     themſelves
    -0.32
    POSITIVE LOGITS
     haven
    1.01
     have
    0.96
     got
    0.84
     want
    0.84
     wondered
    0.83
     cannot
    0.80
     struggled
    0.79
     wanted
    0.79
     havent
    0.79
     merasa
    0.78
    Activations Density 0.315%

    No Known Activations