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. 12779
    Prev
    Next
    INDEX
    Explanations

    conditional statements and arguments regarding exceptions or guarantees

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

    that indicate contrast or contradiction

    np_acts-logits-general · gemini-2.0-flash

    but, however, instead

    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
    akujem
    -0.59
     mergeFrom
    -0.58
    SharedDtor
    -0.56
    InvalidProtocol
    -0.55
    StoreMessageInfo
    -0.54
    onOptions
    -0.54
     Roskov
    -0.53
    <bos>
    -0.53
    UnusedPrivate
    -0.51
    rylic
    -0.51
    POSITIVE LOGITS
     revanche
    1.03
     sí
    1.03
     כן
    1.01
     however
    0.98
    倒是
    0.97
    however
    0.93
     зато
    0.90
    代わりに
    0.86
     viszont
    0.80
    Instead
    0.79
    Activations Density 0.414%

    No Known Activations