© 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. Gemma-2-2B
    3. 22-GEMMASCOPE-TRANSCODER-16K
    4. 10613
    Prev
    Next
    INDEX
    Explanations

    words related to measurements, technology and coding.

    oai_token-act-pair · gemini-2.0-flash

    scientific/technical texts

    np_max-act-logits · gemini-2.0-flash

    s Method used: 1 — reason: max tokens are suffixes ending in "s" — explanation: s

    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_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.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
     للاسماء
    -1.06
     Paglinawan
    -0.98
     AssemblyProduct
    -0.90
    oredCriteria
    -0.90
     Roskov
    -0.89
     الحره
    -0.88
     linkovi
    -0.85
    ⓧ
    -0.81
    Билгалдахарш
    -0.80
     صوتيه
    -0.80
    POSITIVE LOGITS
    ervice
    0.40
    cript
    0.31
    how
    0.31
    oft
    0.31
    erv
    0.30
    cope
    0.30
    ource
    0.29
    hop
    0.29
    noop
    0.29
    cripts
    0.27
    Activations Density 12.401%

    No Known Activations