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. 15-GEMMASCOPE-RES-16K
    6. 4128
    Prev
    Next
    INDEX
    Explanations

    phrases related to simplification and ease of understanding in communication

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_15/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.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Offisielt
    -0.49
    setClass
    -0.47
     nothing
    -0.45
    ัจ
    -0.43
     pourtant
    -0.43
    gios
    -0.42
     rectified
    -0.41
    biskan
    -0.41
    opens
    -0.40
     replaced
    -0.40
    POSITIVE LOGITS
     easier
    1.40
    easier
    1.27
    Easier
    1.15
     Easier
    1.13
    更容易
    0.98
    easily
    0.89
     leichter
    0.89
     easily
    0.87
     quicker
    0.86
     легче
    0.85
    Activations Density 0.348%

    No Known Activations