© 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. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 31-LLAMASCOPE-RES-32K
    11. 4876
    Prev
    Next
    INDEX
    Explanations

    instances of the word "evoke" and its variations, indicating a focus on the theme of eliciting emotions or responses

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    implify
    -0.19
    fea
    -0.15
    aimassage
    -0.15
    ekim
    -0.15
    .gg
    -0.14
    alin
    -0.14
    alion
    -0.14
    arity
    -0.14
    pig
    -0.14
    mai
    -0.14
    POSITIVE LOGITS
    asive
    0.34
    inces
    0.29
    okes
    0.26
    oked
    0.26
    oking
    0.25
    aporation
    0.25
    anes
    0.25
    asions
    0.24
    ildo
    0.24
    ince
    0.24
    Activations Density 0.006%

    No Known Activations