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. 4-GEMMASCOPE-RES-16K
    6. 10934
    Prev
    Next
    INDEX
    Explanations

    instances of the underscore character followed by a word or expression indicating a variable or method in programming contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.4.hook_resid_post
    Hook Layer
    4
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     H
    -0.64
     (
    -0.62
     "
    -0.60
     qui
    -0.59
     h
    -0.59
     ha
    -0.58
     her
    -0.58
    </b>
    -0.58
    ]
    -0.57
     que
    -0.57
    POSITIVE LOGITS
    ._
    2.34
    ]._
    1.69
    .__
    1.45
    )._
    1.41
     pleaſure
    1.30
     myſelf
    1.27
     purpoſe
    1.27
     Reſ
    1.24
     houſe
    1.19
     Inſ
    1.18
    Activations Density 0.095%

    No Known Activations