Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 15-GEMMASCOPE-RES-16K
    6. 2853
    Prev
    Next
    INDEX
    Explanations

    conversational transitions and expressions of digression

    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
    AndEndTag
    -0.90
    +#+#
    -0.87
     Pourtant
    -0.79
     виправивши
    -0.74
     auroit
    -0.73
     hés
    -0.72
    EndInit
    -0.71
    StructEnd
    -0.71
     Monfieur
    -0.70
     Abonnez
    -0.69
    POSITIVE LOGITS
     digress
    0.81
    Anyway
    0.61
     details
    0.60
     Briefly
    0.55
    Back
    0.52
    Details
    0.51
     Anyway
    0.51
    details
    0.50
    Anyways
    0.48
     Back
    0.46
    Activations Density 0.178%

    No Known Activations