© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 8875
    Prev
    Next
    INDEX
    Explanations

    suit

    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
    eration
    -0.10
    媳
    -0.10
    ampton
    -0.10
    æĭĬ
    -0.09
    ä¸įåİ»
    -0.09
     Bounty
    -0.09
    ¤
    -0.09
    byn
    -0.09
    -quarters
    -0.08
    è£Ķ
    -0.08
    POSITIVE LOGITS
    èĩ¾
    0.13
    ties
    0.12
    iful
    0.12
    ably
    0.12
    iveness
    0.10
    ively
    0.10
    cases
    0.10
    lets
    0.10
    èĶ¼
    0.10
    tings
    0.09
    Activations Density 0.011%

    No Known Activations