© 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. 30-TRANSCODER-HP
    7. 96200
    Prev
    Next
    INDEX
    Explanations

    negative situations

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     injecting
    -0.29
    essen
    -0.29
    collapsed
    -0.28
    ovable
    -0.27
    注åħ¥
    -0.26
    æ³½
    -0.25
    practice
    -0.25
    æ¸ł
    -0.25
    .routing
    -0.24
    massage
    -0.24
    POSITIVE LOGITS
     seems
    0.40
     cannot
    0.36
     seem
    0.34
     other
    0.33
     would
    0.31
    å°±ä¸įèĥ½
    0.31
     seemed
    0.31
     exists
    0.30
    éĥ½æĹłæ³ķ
    0.29
     appears
    0.29
    Activations Density 0.889%

    No Known Activations