© 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. 18-TRANSCODER-HP
    7. 89525
    Prev
    Next
    INDEX
    Explanations

    scientific research articles

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    refix
    -0.26
    shaft
    -0.26
    itan
    -0.26
    壤
    -0.25
    FromBody
    -0.25
    ushman
    -0.24
     proceeds
    -0.24
    è¶Ĭå¤ļ
    -0.24
    icals
    -0.24
     suce
    -0.23
    POSITIVE LOGITS
    åįĬå²Ľ
    0.28
    è´µ
    0.27
    CAS
    0.27
    çĴ§
    0.27
    èĶº
    0.26
    ç¾İå¾·
    0.26
     pers
    0.26
    èĢħçļĦ
    0.25
    æ¯ıä¸Ģä½į
    0.25
     stable
    0.25
    Activations Density 9.118%

    No Known Activations