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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 0-GEMMASCOPE-TRANSCODER-16K
    7. 10315
    Prev
    Next
    INDEX
    Explanations

    terms related to law, immigration, and gaming.

    oai_token-act-pair · gemini-2.0-flash

    Rambo

    np_max-act · gemini-2.0-flash

    character names Method used: 1 — because the top activating tokens are proper names

    np_max-act-logits · gpt-5-miniTriggered by @chenshw0109
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_0/width_16k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.0.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    kháu
    -0.73
    ardonnay
    -0.73
    uyasha
    -0.71
    ollett
    -0.68
     judiciales
    -0.65
     Bourgoin
    -0.65
     pères
    -0.65
    OGND
    -0.64
    ]';
    -0.64
    upuncture
    -0.63
    POSITIVE LOGITS
     <<<<<<<<<<<<<<
    0.74
     deadliest
    0.66
     talking
    0.64
     luxo
    0.63
     Pfal
    0.61
    breakable
    0.60
     resilient
    0.60
     Advanced
    0.60
     snapping
    0.59
     immensely
    0.59
    Activations Density 8.976%

    No Known Activations