© 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. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 22402
    Prev
    Next
    INDEX
    Explanations

    mentions of flaws and design aspects in performance-related contexts

    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
    amam
    -0.16
    asures
    -0.14
    uby
    -0.13
     blas
    -0.13
    orphic
    -0.13
    -devel
    -0.13
    ander
    -0.13
    oltage
    -0.13
    jem
    -0.13
     delayed
    -0.13
    POSITIVE LOGITS
     faults
    0.53
     fault
    0.53
     weaknesses
    0.51
     flaw
    0.51
     flaws
    0.49
     Fault
    0.49
     weakness
    0.47
    fault
    0.45
     defects
    0.45
     Weak
    0.45
    Activations Density 0.070%

    No Known Activations