© 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. 108010
    Prev
    Next
    INDEX
    Explanations

    is

    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
    äºĨä¸Ģ份
    -0.29
     ÑĦакÑĤ
    -0.28
    麻çĹ¹
    -0.27
    (act
    -0.27
    ä¸į说
    -0.27
    æŃ»åİ»
    -0.26
    fact
    -0.25
    फ
    -0.25
    äºĨå¤ļå°ij
    -0.24
    ä¸Ģåı¥
    -0.23
    POSITIVE LOGITS
    å°¤åħ¶
    0.38
    å°¤åħ¶æĺ¯
    0.35
    å°¤åħ¶æĺ¯åľ¨
    0.34
     especially
    0.34
     exempl
    0.31
    æĹ¢æľī
    0.29
    çī¹åĪ«æĺ¯åľ¨
    0.28
     besonders
    0.28
    genre
    0.28
     particularly
    0.28
    Activations Density 1.079%

    No Known Activations