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. 26-LLAMASCOPE-RES-32K
    6. 28012
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "celebrate" and related terms, highlighting themes of celebration and significant events

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.26.hook_resid_post
    Hook Layer
    26
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     celebration
    -0.33
     celebrate
    -0.31
     celebr
    -0.30
     Celebration
    -0.29
    Celebr
    -0.29
     celebrating
    -0.28
     Celebr
    -0.28
     celebrates
    -0.28
    cele
    -0.27
     celebrations
    -0.26
    POSITIVE LOGITS
    atory
    0.19
    iber
    0.16
    zik
    0.16
     differences
    0.15
     Differences
    0.15
    -worthy
    0.15
    á¹
    0.15
    zza
    0.14
    Together
    0.14
     being
    0.14
    Activations Density 0.027%

    No Known Activations