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. 21-GEMMASCOPE-RES-16K
    6. 7383
    Prev
    Next
    INDEX
    Explanations

    discussions around societal issues, particularly related to accusations and how they impact individuals

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.hook_resid_post
    Hook Layer
    21
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     poffe
    -1.04
     purpoſe
    -1.04
     pleaſure
    -1.03
     deſt
    -0.97
     houſe
    -0.96
     fubject
    -0.95
     ſever
    -0.94
     tranſ
    -0.93
     neceff
    -0.93
     himſelf
    -0.92
    POSITIVE LOGITS
     whatnot
    0.79
     stuff
    0.75
     things
    0.69
     maybe
    0.68
     thingy
    0.66
    ,
    0.64
     et
    0.60
     doings
    0.60
     Maybe
    0.59
    maybe
    0.58
    Activations Density 0.407%

    No Known Activations