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-2-2B
    4. Residual Stream - 16k
    5. 16-GEMMASCOPE-RES-16K
    6. 16124
    Prev
    Next
    INDEX
    Explanations

    special character sequences and string formatting patterns in coding and markup languages

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

    Characters often preceding numbers or code

    np_acts-logits-general · gemini-2.0-flash

    special characters and programming syntax

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 16-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    AsUp
    -0.94
    djangoproject
    -0.84
    fillType
    -0.73
    mybatisplus
    -0.70
    webElement
    -0.67
    nocache
    -0.67
    Przypisy
    -0.66
    msgSender
    -0.65
    typeorm
    -0.64
    RepeatedField
    -0.62
    POSITIVE LOGITS
     raiſ
    0.89
     ſen
    0.81
    ="@+
    0.77
     juſt
    0.76
     $("<
    0.76
     pleaſure
    0.74
     ſon
    0.74
     ſer
    0.74
    ]<<"
    0.73
    __*/
    0.71
    Activations Density 0.264%

    No Known Activations