© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 28952
    Prev
    Next
    INDEX
    Explanations

    phrases related to sports plays and game actions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Juliushanhanhan/llama-3-8b-it-res/blocks.25.hook_resid_post
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.25.hook_resid_post
    Hook Layer
    25
    Architecture
    gated
    Context Size
    1,024
    Dataset
    Juliushanhanhan/openwebtext-1b-llama3-tokenized-cxt-1024
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    zsche
    -0.17
    adata
    -0.16
    ç½
    -0.15
    ller
    -0.14
     casts
    -0.14
    153
    -0.14
    ENCH
    -0.14
    ecast
    -0.14
     Arbitrary
    -0.14
    istra
    -0.14
    POSITIVE LOGITS
     inbound
    0.22
     overtime
    0.19
     attempts
    0.18
     POSS
    0.17
    attempt
    0.17
     tying
    0.17
     attempt
    0.17
     attempted
    0.17
     ë§Īì§Ģë§ī
    0.17
     timeouts
    0.16
    Activations Density 0.011%

    No Known Activations