© 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. 55137
    Prev
    Next
    INDEX
    Explanations

    from one to another

    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
    è¶Ĭå¤ļ
    -0.10
    PropertyParams
    -0.10
    sten
    -0.10
    italic
    -0.09
    åħ¶å®ĥéĹ®é¢ĺ
    -0.09
    _SCOPE
    -0.09
    çĻ¾å®¶
    -0.09
    _procs
    -0.09
    å¼ĥ
    -0.09
    clang
    -0.08
    POSITIVE LOGITS
    æİ¥åĪ°
    0.09
    åł¡åŀĴ
    0.09
    iator
    0.08
    ors
    0.08
    çĸı导
    0.08
    ively
    0.08
    è§ģåĪ°
    0.08
    大å±Ģ
    0.08
    å±ķçİ°
    0.08
     Debbie
    0.07
    Activations Density 0.045%

    No Known Activations