© 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. 108951
    Prev
    Next
    INDEX
    Explanations

    Intensifies adjectives/frustration

    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
    ophon
    -0.27
    bright
    -0.27
    usch
    -0.26
    æģŃæķ¬
    -0.26
    bone
    -0.25
     Choir
    -0.24
    ursal
    -0.24
     diffuse
    -0.24
    UMB
    -0.23
     cheerful
    -0.23
    POSITIVE LOGITS
    itzer
    0.29
     stre
    0.28
    ieri
    0.27
     Mon
    0.25
    çİĩ为
    0.25
    Ñģол
    0.25
    еÑĢи
    0.25
    ieres
    0.24
    unc
    0.24
    arte
    0.24
    Activations Density 1.153%

    No Known Activations