© 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. 108478
    Prev
    Next
    INDEX
    Explanations

    cache

    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
    æĬķéĻį
    -0.30
    åĽ´
    -0.27
    讼
    -0.26
    éĻį
    -0.26
    trait
    -0.25
    éĹ¹
    -0.25
     pulp
    -0.25
    íĢĢ
    -0.25
    li
    -0.25
     spat
    -0.24
    POSITIVE LOGITS
     cache
    1.09
    ç¼ĵ
    1.06
     caching
    1.03
     cached
    1.02
    cache
    0.98
     Cache
    0.98
     caches
    0.97
    _cache
    0.94
    -cache
    0.94
    	cache
    0.93
    Activations Density 1.139%

    No Known Activations