© 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. 11475
    Prev
    Next
    INDEX
    Explanations

    escape or out

    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¿ľç¨ĭ
    -0.10
     remote
    -0.10
    Remote
    -0.09
    remote
    -0.09
     Remote
    -0.09
    _remote
    -0.09
    é¦ĸåıij
    -0.08
    èµī
    -0.08
    æ¹ĥ
    -0.08
    .remote
    -0.08
    POSITIVE LOGITS
    è·³åĩº
    0.22
    æijĨèĦ±
    0.20
    éĢĥèĦ±
    0.20
    éĢĥ离
    0.20
     escape
    0.18
    æĬľãģij
    0.18
    èĦ±ç¦»
    0.17
    èµ°åĩºæĿ¥
    0.16
     Escape
    0.15
     escapes
    0.15
    Activations Density 0.101%

    No Known Activations