© 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. 35905
    Prev
    Next
    INDEX
    Explanations

    preferred approach or configuration

    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
    çĺª
    -0.09
    ToOne
    -0.09
    èĶ¼
    -0.08
    aware
    -0.08
    åĦ¿
    -0.08
    åħĴ
    -0.08
    .ie
    -0.08
    orial
    -0.08
    dag
    -0.08
    dw
    -0.08
    POSITIVE LOGITS
    entially
    0.12
    çļĦåľ°ä½į
    0.10
    ennen
    0.10
    殿ä¸ĭ
    0.09
    åľ°ä½į
    0.09
    æ»ĭåij³
    0.09
    ç®ĢåİĨ
    0.09
    æĿ¼
    0.09
    å¤ļãģı
    0.09
    è¿Ļç§įæĸ¹æ³ķ
    0.09
    Activations Density 0.028%

    No Known Activations