© 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
    2. Gemma-3-27B-IT
    3. 15-GEMMASCOPE-2-TRANSCODER-262K
    4. 14663
    Prev
    Next
    INDEX
    Explanations

    **Explaining Code/Technical Descriptions**I'm synthesizing the information to deliver a precise explanation of the neuron's behavior. I will focus on identifying the core theme from the positive logits, activating texts, and token sequences, ensuring the output is a concise phrase that captures its function in a technical or coding context.I'm synthesizing the provided information to identify the underlying pattern and generate a concise explanation for the neuron's behavior. The focus will be on the semantic meaning derived from the positive

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_15_width_262k_l0_small_affine
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     hoera
    0.50
    чә
    0.49
    OfString
    0.48
     ఆధ్యాత్మిక
    0.47
    il
    0.47
    FlashValue
    0.46
     என்ன
    0.46
    GoString
    0.46
    ческую
    0.46
    OfDeath
    0.46
    POSITIVE LOGITS
     acquired
    0.51
    veloc
    0.50
     Ná
    0.50
    P
    0.49
    alka
    0.48
    olos
    0.47
     inac
    0.45
    winfx
    0.45
     noviembre
    0.45
     SPDX
    0.44
    Activations Density 0.001%

    No Known Activations