© 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. 31-TRANSCODER-HP
    7. 18979
    Prev
    Next
    INDEX
    Explanations

    "Enough" or "Needless to say"

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åĭ¤åĬ³
    -0.30
    degrees
    -0.28
    ows
    -0.28
    OLER
    -0.26
    laÅŁ
    -0.26
    uti
    -0.26
    bih
    -0.26
     degrees
    -0.25
    ativas
    -0.25
    osate
    -0.25
    POSITIVE LOGITS
    -nav
    0.26
    çļĦåIJĮæŶ
    0.26
    éĩįæķ´
    0.25
    æ··
    0.25
     messing
    0.25
    rell
    0.25
    å®īåįĵ
    0.24
     Already
    0.24
    èĺ¸
    0.24
    没
    0.24
    Activations Density 0.005%

    No Known Activations