© 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. 20-LLAMASCOPE-RES-32K
    11. 12996
    Prev
    Next
    INDEX
    Explanations

    phrases indicating conditions or requirements for success or effectiveness

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    zet
    -0.14
    #
    -0.13
    ista
    -0.13
    iqueta
    -0.13
    quential
    -0.13
    zte
    -0.13
     rast
    -0.13
    ØŃÙĬØ©
    -0.12
    hek
    -0.12
    Ä±ÅŁÄ±k
    -0.12
    POSITIVE LOGITS
     how
    0.68
    how
    0.54
     what
    0.51
     why
    0.44
     whether
    0.42
     cómo
    0.41
    what
    0.41
     How
    0.39
    å¦Ĥä½ķ
    0.38
     where
    0.37
    Activations Density 0.157%

    No Known Activations