© 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. 62708
    Prev
    Next
    INDEX
    Explanations

    tunnel or tunneling concepts

    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
    enance
    -0.10
    fully
    -0.08
    ç»
    -0.08
    nts
    -0.08
    yp
    -0.08
    enstein
    -0.08
    ynchronously
    -0.08
     trous
    -0.08
    ternal
    -0.08
    aceous
    -0.08
    POSITIVE LOGITS
    led
    0.11
    trie
    0.10
    blick
    0.10
    maid
    0.09
    rand
    0.09
    abor
    0.09
    mad
    0.09
    å²ĺ
    0.09
    æ·±å¤Ħ
    0.09
    IGHT
    0.09
    Activations Density 0.020%

    No Known Activations