© 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 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. 28292
    Prev
    Next
    INDEX
    Explanations

    patterns or structures resembling visual or abstract representations

    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
    ucer
    -0.15
    \Php
    -0.14
    ï¸ı
    -0.13
    etag
    -0.13
    awi
    -0.13
    onder
    -0.13
    âĢķâĢķâĢķâĢķ
    -0.13
    acle
    -0.13
    ecz
    -0.13
    Lexer
    -0.13
    POSITIVE LOGITS
    ẽ
    0.15
    iry
    0.15
    .ImageAlign
    0.15
     Portal
    0.15
     counsel
    0.14
    ëĭ
    0.13
    Portal
    0.13
     AssemblyVersion
    0.13
    ¢åįķ
    0.13
    à¸Ħà¸Ļ
    0.13
    Activations Density 0.012%

    No Known Activations