© 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. 16-LLAMASCOPE-RES-32K
    11. 17972
    Prev
    Next
    INDEX
    Explanations

    proper nouns, particularly names of people and entities related to the film and political industry

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ÃĸL
    -0.17
    ÅĻi
    -0.16
    ONGO
    -0.15
    auga
    -0.15
    quoise
    -0.14
    esh
    -0.14
    èĢ
    -0.14
    Ñģион
    -0.14
    angen
    -0.14
    Ahead
    -0.14
    POSITIVE LOGITS
    æ¤
    0.14
    deck
    0.14
    saldo
    0.14
    illow
    0.13
     lesbi
    0.13
    ivec
    0.13
     eens
    0.13
    ertools
    0.12
     Turk
    0.12
    anyahu
    0.12
    Activations Density 0.197%

    No Known Activations