© 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. 16-LLAMASCOPE-RES-32K
    11. 386
    Prev
    Next
    INDEX
    Explanations

    instances of the phrase "of" followed by various expressions or qualifiers

    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
    .FontStyle
    -0.18
    orry
    -0.16
    resses
    -0.15
    ment
    -0.14
    arians
    -0.14
    eyim
    -0.14
    suit
    -0.14
    ãĥ¡ãĥ³ãĥĪ
    -0.14
    clf
    -0.14
    658
    -0.14
    POSITIVE LOGITS
    ataka
    0.17
     lab
    0.14
    malink
    0.14
     discrepan
    0.14
    andest
    0.14
     tun
    0.14
     yet
    0.13
     Breed
    0.13
    yet
    0.13
     barred
    0.13
    Activations Density 0.019%

    No Known Activations