© 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. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 58560
    Prev
    Next
    INDEX
    Explanations

    jump + preposition/adverb

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    èĺħ
    -0.10
    ä¸įæĺİ
    -0.09
    haus
    -0.09
    icens
    -0.09
    pgsql
    -0.09
    etas
    -0.08
    çĽ¼
    -0.08
    afi
    -0.08
    _DSP
    -0.08
    opus
    -0.08
    POSITIVE LOGITS
    äºĨä¸įå°ij
    0.09
    çŀĮ
    0.09
    ey
    0.09
    éĩľ
    0.09
    ç§ij室
    0.09
    äºĨå¾Īå¤ļ
    0.09
    åĴĮå®ĮåĸĦ
    0.08
    ä»ħ次äºİ
    0.08
    ç»ĨåĮĸ
    0.08
    èĺ¸
    0.08
    Activations Density 0.008%

    No Known Activations