© 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. 15-LLAMASCOPE-RES-32K
    11. 308
    Prev
    Next
    INDEX
    Explanations

    negative symbols or terms related to coding and command-line operations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    command-line flags and options (tokens starting with a hyphen/dash).

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    command-line option flags prefixed with dashes in shell or build commands.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 15-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L15R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ascus
    -0.19
    rouw
    -0.18
    _exempt
    -0.14
    esti
    -0.14
    riage
    -0.14
    rière
    -0.14
    vably
    -0.14
    edException
    -0.14
    ir
    -0.14
    indow
    -0.13
    POSITIVE LOGITS
     force
    0.17
     quiet
    0.15
    force
    0.15
    ller
    0.14
    Quiet
    0.14
    quiet
    0.14
    eron
    0.14
    une
    0.14
    azio
    0.14
    q
    0.13
    Activations Density 0.013%

    No Known Activations