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. 18-GEMMASCOPE-RES-16K
    6. 14533
    Prev
    Next
    INDEX
    Explanations

    structures in software code related to logging and error handling

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

    Punctuation in code or text files

    np_acts-logits-general · gemini-2.0-flash

    programming, code, french, german, russian

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     للمعارف
    -0.71
     بيها
    -0.67
    ConstraintMaker
    -0.63
     كومونز
    -0.57
    إنه
    -0.55
     Efq
    -0.54
     Wikimedijinoj
    -0.53
    Filmografia
    -0.53
    +#+
    -0.52
    KommentareTeilen
    -0.52
    POSITIVE LOGITS
    principalTable
    0.58
     незавершена
    0.53
    createComponent
    0.53
    RunAsync
    0.52
     dringend
    0.51
    ExecuteReader
    0.50
    '}';
    0.50
     Référence
    0.49
    leby
    0.49
    wendungs
    0.48
    Activations Density 0.259%

    No Known Activations