© 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. 31-LLAMASCOPE-RES-32K
    11. 22102
    Prev
    Next
    INDEX
    Explanations

    references to self in programming or testing contexts

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Negative Logits
    .updateDynamic
    -0.19
    .GridColumn
    -0.14
     authorized
    -0.14
     Authorized
    -0.13
     one
    -0.13
     Davies
    -0.13
    ष
    -0.13
    igne
    -0.13
    $LANG
    -0.13
     Chap
    -0.13
    POSITIVE LOGITS
    .assertEqual
    0.40
    .assertTrue
    0.38
    .assertEquals
    0.36
    .assertFalse
    0.33
     expected
    0.31
     expectation
    0.31
     assertEquals
    0.31
     expect
    0.31
     assert
    0.31
    .assertNot
    0.30
    Activations Density 0.005%

    No Known Activations