© 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. 12060
    Prev
    Next
    INDEX
    Explanations

    the word "group" and words related to tools or instruments

    oai_token-act-pair · gemini-2.0-flash

    group

    np_max-act · gemini-2.0-flash

    say "prepositions or commas" Method used: 2 — because following tokens are common prepositions or punctuation

    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
    httphttps
    -0.83
    DockStyle
    -0.81
     Signalez
    -0.72
    OGND
    -0.72
    #
    -0.69
    /**
    -0.68
     debout
    -0.68
    +#+#
    -0.68
     Gost
    -0.65
     roul
    -0.65
    POSITIVE LOGITS
    otesi
    0.64
    xically
    0.63
    refour
    0.59
    ndice
    0.55
    PutMapping
    0.52
    izophren
    0.52
    pecies
    0.51
    rentice
    0.50
     supposed
    0.50
    Zeneca
    0.50
    Activations Density 0.642%

    No Known Activations