© 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. 23-LLAMASCOPE-RES-32K
    11. 824
    Prev
    Next
    INDEX
    Explanations

    instances of the word "eight" and its variations, indicating a focus on the number eight and its contextual significance

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.23.hook_resid_post
    Hook Layer
    23
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    itional
    -0.19
    ร
    -0.19
    fter
    -0.17
    etal
    -0.16
    atable
    -0.15
    ÏĥειÏĤ
    -0.15
    west
    -0.15
    conti
    -0.15
    ibrate
    -0.15
    792
    -0.15
    POSITIVE LOGITS
    eenth
    0.38
    een
    0.28
    ãģ¤ãģ®
    0.22
    yi
    0.19
    ies
    0.19
    ï¸ı
    0.18
    erez
    0.18
    *sizeof
    0.18
    ting
    0.16
    enment
    0.16
    Activations Density 0.085%

    No Known Activations