© 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. 108612
    Prev
    Next
    INDEX
    Explanations

    IQ

    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
    ryo
    -0.32
     gr
    -0.28
    ÑģÑĤвеннÑĭй
    -0.27
     cou
    -0.26
    èģļ
    -0.24
    im
    -0.24
    ãĥĩãĤ£ãĥ³ãĤ°
    -0.24
     ordin
    -0.23
     previous
    -0.23
    液
    -0.23
    POSITIVE LOGITS
    ris
    0.24
    gart
    0.24
    è
    0.24
    lesia
    0.23
    akedown
    0.23
    atta
    0.23
     Reggie
    0.23
    éłħ
    0.23
    éĽĨä½ĵ
    0.23
     sẵn
    0.22
    Activations Density 0.108%

    No Known Activations