© 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. 6-GEMMASCOPE-TRANSCODER-16K
    7. 13058
    Prev
    Next
    INDEX
    Explanations

    words related to abstract philosophical concepts as well as hardship ("odities", "vague ideas", "poor")

    oai_token-act-pair · gemini-2.0-flash

    Strange and hard to understand

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_6/width_16k/average_l0_95
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.6.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
     üzere
    -0.51
     Himo
    -0.51
    dataclass
    -0.49
     tagli
    -0.46
    formik
    -0.44
     geht
    -0.43
     بيها
    -0.42
     graduated
    -0.42
     gerne
    -0.41
    Дей
    -0.41
    POSITIVE LOGITS
    utilisons
    0.70
    hastly
    0.65
     AssemblyCompany
    0.65
    archiviato
    0.64
    phazard
    0.63
    itism
    0.63
    celess
    0.62
    expectedResult
    0.62
    minecraftforge
    0.62
    MergeFrom
    0.61
    Activations Density 3.823%

    No Known Activations