© 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. 21-LLAMASCOPE-RES-32K
    11. 8870
    Prev
    Next
    INDEX
    Explanations

    programming-related function calls and variable assignment

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.21.hook_resid_post
    Hook Layer
    21
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Down
    -0.15
    ually
    -0.14
    ORMAL
    -0.14
    inand
    -0.14
    anzi
    -0.14
    /Common
    -0.14
    rant
    -0.14
    uzzi
    -0.14
    å¢
    -0.13
    adem
    -0.13
    POSITIVE LOGITS
    tle
    0.28
    sockopt
    0.24
    /get
    0.22
     aside
    0.21
    =set
    0.21
    uptools
    0.20
    metatable
    0.19
    jmp
    0.18
    embro
    0.18
    elah
    0.18
    Activations Density 0.064%

    No Known Activations