© 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. 33383
    Prev
    Next
    INDEX
    Explanations

    evaluations and assessments of individuals, particularly in a political context

    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
     NSK
    -0.17
    erland
    -0.16
    ruba
    -0.15
     Seeder
    -0.15
    TriState
    -0.15
    ìĺ¥
    -0.15
    OffsetTable
    -0.15
    ABCDEFGHIJKLMNOP
    -0.14
     Tradable
    -0.14
    çĢ
    -0.14
    POSITIVE LOGITS
     dumb
    0.48
     stupid
    0.45
     stupidity
    0.45
     idi
    0.45
     dummy
    0.43
     Dummy
    0.42
     idiot
    0.42
    Dummy
    0.40
     Id
    0.40
    dummy
    0.39
    Activations Density 0.061%

    No Known Activations