© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 14-LLAMASCOPE-RES-32K
    11. 18539
    Prev
    Next
    INDEX
    Explanations

    concepts related to structural changes, especially in a systematic and reformative context

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.14.hook_resid_post
    Hook Layer
    14
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Negative Logits
    δα
    -0.17
    warts
    -0.17
    /run
    -0.15
    ross
    -0.15
     seins
    -0.14
    lesh
    -0.14
    /extensions
    -0.14
    dag
    -0.14
    ocks
    -0.14
     sublic
    -0.14
    POSITIVE LOGITS
     Improvement
    0.24
     improvement
    0.22
     improve
    0.22
     change
    0.20
     Improve
    0.20
     improvements
    0.20
     improves
    0.20
    change
    0.19
     Impro
    0.19
     improving
    0.19
    Activations Density 0.473%

    No Known Activations