© 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. 25-TRANSCODER-HP
    7. 107883
    Prev
    Next
    INDEX
    Explanations

    Negative words

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    infinity
    -0.26
    èĭ
    -0.26
    纳米
    -0.25
    æºľ
    -0.25
    space
    -0.25
    å¾®è§Ĥ
    -0.25
     honorable
    -0.25
    å®ıè§Ĥ
    -0.25
    鸳
    -0.24
    vic
    -0.24
    POSITIVE LOGITS
     fully
    0.32
    æĺ¯æ²¡æľī
    0.31
    iert
    0.30
    agedList
    0.29
    pleted
    0.28
    iated
    0.27
     minim
    0.27
    Fully
    0.27
     no
    0.27
     without
    0.27
    Activations Density 2.768%

    No Known Activations