© 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. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 26-LLAMASCOPE-RES-32K
    11. 9917
    Prev
    Next
    INDEX
    Explanations

    words related to manipulation or influence

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.26.hook_resid_post
    Hook Layer
    26
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    mites
    -0.17
    lien
    -0.15
    ouri
    -0.15
    ereum
    -0.14
    -cn
    -0.14
    embers
    -0.14
     onError
    -0.14
     Orn
    -0.14
    Console
    -0.14
     Buch
    -0.14
    POSITIVE LOGITS
     adipiscing
    0.20
    edly
    0.17
    uent
    0.17
    radi
    0.16
    orde
    0.16
    away
    0.16
    ÏĥÏĦαν
    0.16
     adipisicing
    0.16
    etur
    0.16
    isseur
    0.15
    Activations Density 0.032%

    No Known Activations