© 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. 10782
    Prev
    Next
    INDEX
    Explanations

    swap or exchange

    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
    NECT
    -0.12
    éģı
    -0.10
    abus
    -0.10
    çĶ©
    -0.09
    TIMER
    -0.09
    罩
    -0.09
    åĽŀé¦Ī
    -0.09
    AVED
    -0.09
    esimal
    -0.09
    avax
    -0.09
    POSITIVE LOGITS
     exchange
    0.16
    交æį¢
    0.15
    exchange
    0.14
    .exchange
    0.13
     swap
    0.13
     Exchange
    0.12
    swap
    0.12
     exchanges
    0.12
    .Exchange
    0.12
    æį¢äºĨ
    0.11
    Activations Density 0.086%

    No Known Activations