© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    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. 2064
    Prev
    Next
    INDEX
    Explanations

    dash or hyphen symbols in textual formatting

    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
     Efq
    -0.72
     Jefus
    -0.68
    omiast
    -0.68
     Houſe
    -0.66
    ſelf
    -0.65
     Majefty
    -0.64
     cetera
    -0.64
    ſelves
    -0.63
     stb
    -0.62
     TAMBIÉN
    -0.62
    POSITIVE LOGITS
     although
    0.92
     but
    0.85
     albeit
    0.83
     namely
    0.81
     including
    0.80
    although
    0.79
     especially
    0.76
     perhaps
    0.76
    except
    0.76
     despite
    0.76
    Activations Density 0.114%

    No Known Activations