© 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. 5427
    Prev
    Next
    INDEX
    Explanations

    rail or rails

    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
    ente
    -0.10
    naire
    -0.10
    æĹĦ
    -0.09
    ëĿ¼ëıĦ
    -0.08
    çĶ©
    -0.08
    åĭĥåĭĥ
    -0.08
    ality
    -0.08
    .NULL
    -0.08
    åĽ±
    -0.08
    ä¸ļæĢģ
    -0.08
    POSITIVE LOGITS
    ings
    0.11
    mates
    0.10
    yard
    0.10
    roads
    0.10
    ments
    0.10
    uet
    0.10
    bud
    0.10
    UMENT
    0.10
    ways
    0.09
    utes
    0.09
    Activations Density 0.026%

    No Known Activations