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. 21-LLAMASCOPE-RES-32K
    6. 9980
    Prev
    Next
    INDEX
    Explanations

    phrases related to opportunity and potential for growth

    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
    ëŁ¼
    -0.16
    evin
    -0.15
    reon
    -0.14
    chg
    -0.14
    ioni
    -0.14
    лÑĥÑĪ
    -0.14
    isson
    -0.14
    rema
    -0.14
    ubble
    -0.13
    kla
    -0.13
    POSITIVE LOGITS
     aprove
    0.55
     exploit
    0.53
     exploitation
    0.49
     exploited
    0.46
     capital
    0.44
     capitalize
    0.44
     exploiting
    0.42
    åĪ©ç͍
    0.42
     exploits
    0.41
    explo
    0.39
    Activations Density 0.297%

    No Known Activations