© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 20182
    Prev
    Next
    INDEX
    Explanations

    record or remember past actions

    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
    (mx
    -0.09
    èĬĴ
    -0.08
    =ax
    -0.08
    èµ°
    -0.08
    RelativeTo
    -0.08
     недо
    -0.08
    è±³
    -0.08
    é¦ĸ
    -0.08
     ble
    -0.08
    anity
    -0.08
    POSITIVE LOGITS
    计è¾ĥ
    0.13
    å¾Ģäºĭ
    0.12
    è®°å¿ĨåĬĽ
    0.12
    追究
    0.12
    è¨ĺ
    0.12
    æŧ
    0.11
    çĿļ
    0.11
    è®°
    0.11
    éĤ£æ¬¡
    0.10
    /details
    0.10
    Activations Density 0.036%

    No Known Activations