© 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. 22-TRANSCODER-HP
    7. 98483
    Prev
    Next
    INDEX
    Explanations

    code-related characters

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    иÑĦикаÑĨи
    -0.25
     Mic
    -0.24
    èĥļèĥİ
    -0.24
    leans
    -0.23
    Mic
    -0.23
    ril
    -0.23
    fried
    -0.23
    RIC
    -0.23
     BehaviorSubject
    -0.23
     lạnh
    -0.23
    POSITIVE LOGITS
    ä¸Ģç»ı
    0.27
     adul
    0.27
    vell
    0.25
    @Test
    0.25
     hon
    0.24
    ä¸Ģéģį
    0.23
    çĢļ
    0.23
     salute
    0.23
     prag
    0.23
    CharacterSet
    0.23
    Activations Density 0.083%

    No Known Activations