© 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. 50966
    Prev
    Next
    INDEX
    Explanations

    filing cabinet or file court documents

    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
    itat
    -0.09
    lops
    -0.09
    çıij
    -0.09
    AGO
    -0.08
    Caller
    -0.08
    .Maximum
    -0.08
    çIJĥåľº
    -0.08
    ibre
    -0.08
    ä¼łå¯¼
    -0.08
    efs
    -0.08
    POSITIVE LOGITS
    æĮĩçĶ²
    0.10
    rim
    0.09
    ogn
    0.09
    夹
    0.09
    å°¼æĸ¯
    0.09
    åĨĻåŃĹ
    0.09
    (File
    0.09
    anchors
    0.08
    -resources
    0.08
    害
    0.08
    Activations Density 0.170%

    No Known Activations