© 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. 33425
    Prev
    Next
    INDEX
    Explanations

    pit, hole, trench, depth

    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
    檬
    -0.09
    çľ·
    -0.08
    rams
    -0.08
    _nil
    -0.08
    æĭĤ
    -0.08
    ply
    -0.08
    /pkg
    -0.08
    è¹ĩ
    -0.08
    ays
    -0.08
    bart
    -0.08
    POSITIVE LOGITS
    å¾Īæ·±
    0.17
    深度
    0.15
    槽
    0.14
    åĬłæ·±
    0.14
     depths
    0.13
    deep
    0.13
    æĽ´æ·±
    0.13
    æ·±
    0.13
    åĿij
    0.12
    å£ij
    0.12
    Activations Density 0.131%

    No Known Activations