© 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. 22978
    Prev
    Next
    INDEX
    Explanations

    references to evidence and the concept of proof, particularly in relation to alien encounters and related research

    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
    oga
    -0.15
     sad
    -0.14
     Sad
    -0.14
    ana
    -0.14
    Sad
    -0.14
     tort
    -0.14
    æļ´
    -0.14
     Pain
    -0.14
    âĶIJ
    -0.13
     Barbar
    -0.13
    POSITIVE LOGITS
     sightings
    0.32
     reports
    0.28
     UFO
    0.26
     phenomenon
    0.25
     strange
    0.24
     phenomena
    0.23
     sighting
    0.23
     reported
    0.23
     paranormal
    0.23
     inexp
    0.23
    Activations Density 0.075%

    No Known Activations