© 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. 18315
    Prev
    Next
    INDEX
    Explanations

    multiples, multipliers, and increases

    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
    thic
    -0.09
    è¹²
    -0.09
    ÑģÑĤÑĢ
    -0.08
    æ¤ĭ
    -0.08
    éºĹ
    -0.08
    æĩ¿
    -0.08
    argas
    -0.08
    触åıĬ
    -0.08
     spoilers
    -0.08
    èģĶ
    -0.08
    POSITIVE LOGITS
    åĢį
    0.21
    åĬłåĢį
    0.17
     gấp
    0.15
     doubled
    0.15
     doubling
    0.13
    ä¸ĢåįĬ
    0.13
     twice
    0.12
     dou
    0.11
    double
    0.10
     Twice
    0.10
    Activations Density 0.067%

    No Known Activations