© 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. Qwen3-4B
    5. 33-TRANSCODER-HP
    7. 52995
    Prev
    Next
    INDEX
    Explanations

    on

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     account
    -0.39
     board
    -0.31
    account
    -0.30
     Account
    -0.30
    å¸IJ
    -0.30
     BOARD
    -0.28
    STANCE
    -0.28
    ä¸ĭä¸Ģ个
    -0.28
    caught
    -0.27
    _account
    -0.27
    POSITIVE LOGITS
    èĤ²
    0.27
    çĽ²
    0.27
    -piece
    0.26
     Revision
    0.26
    æĢĿç»´
    0.25
     literacy
    0.25
    âĦĵ
    0.25
     tvb
    0.25
    estre
    0.25
    æµıè§ĪåĻ¨
    0.25
    Activations Density 0.010%

    No Known Activations