© 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. 108761
    Prev
    Next
    INDEX
    Explanations

    on/of

    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
    订éĺħ
    -0.27
     equip
    -0.27
    è§Ĥæµĭ
    -0.26
    驽
    -0.26
     explicit
    -0.25
    åIJĦ个
    -0.25
    patches
    -0.24
    wares
    -0.24
    尾巴
    -0.24
    çĮ·
    -0.24
    POSITIVE LOGITS
     climbing
    0.29
     both
    0.28
    stown
    0.27
    åħ¶ä¸Ńä¹ĭä¸Ģ
    0.27
    æĹ¢
    0.27
     Aspen
    0.26
    Atl
    0.25
    arden
    0.25
     viable
    0.25
     Among
    0.24
    Activations Density 0.108%

    No Known Activations