© 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. 35-TRANSCODER-HP
    7. 39636
    Prev
    Next
    INDEX
    Explanations

    code and references

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_35.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.35.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    رة
    -0.28
     Cunning
    -0.28
    лÑĮн
    -0.27
    edList
    -0.27
    edException
    -0.26
    ania
    -0.25
    alement
    -0.25
    ание
    -0.25
    lichkeit
    -0.25
    èĢĮåľ¨
    -0.25
    POSITIVE LOGITS
    âģ
    0.31
     bunch
    0.29
     discard
    0.27
     rationale
    0.26
     relinqu
    0.26
    δ
    0.26
    âĨĴ
    0.26
    ^
    0.25
     retain
    0.25
    <|endoftext|>
    0.25
    Activations Density 73.775%

    No Known Activations