© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 3-LLAMASCOPE-RES-32K
    11. 470
    Prev
    Next
    INDEX
    Explanations

    variations of the word "Fab" and its associated forms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 3-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L3R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.3.hook_resid_post
    Hook Layer
    3
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     DataSource
    -0.16
    ses
    -0.16
    ogh
    -0.15
    eldom
    -0.14
    ábado
    -0.14
    ivas
    -0.14
    319
    -0.14
    sume
    -0.13
    uang
    -0.13
    earable
    -0.13
    POSITIVE LOGITS
    ROTO
    0.18
    oucher
    0.16
    OLON
    0.16
    ola
    0.15
    ulous
    0.14
    lsi
    0.14
    CALE
    0.14
    ilit
    0.14
     Caldwell
    0.14
    odia
    0.13
    Activations Density 0.012%

    No Known Activations