Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    3. Llama3.1-8B (Base)
    4. Residual Stream
    5. 30-LLAMASCOPE-RES-32K
    6. 17240
    Prev
    Next
    INDEX
    Explanations

    references to luxury brands and their cultural significance in China

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.30.hook_resid_post
    Hook Layer
    30
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ãĤ¤ãĥĪ
    -0.17
    hd
    -0.14
     Salah
    -0.14
     æĬķ稿
    -0.13
    ãģĵãĤĵãģ«
    -0.13
    emy
    -0.13
    ÐIJÑĢÑħÑĸв
    -0.13
    rxjs
    -0.12
     tamil
    -0.12
    »
    -0.12
    POSITIVE LOGITS
     Chinese
    0.60
     China
    0.59
    China
    0.50
     Beijing
    0.50
    Chinese
    0.49
     chinese
    0.45
     Shanghai
    0.44
     china
    0.42
     Guang
    0.41
     Mandarin
    0.41
    Activations Density 0.627%

    No Known Activations