© 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. 29-LLAMASCOPE-RES-32K
    11. 12190
    Prev
    Next
    INDEX
    Explanations

    instances of the word "find" and its variations, indicating a focus on searching for identity, purpose, or belonging

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.29.hook_resid_post
    Hook Layer
    29
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ALAR
    -0.14
    зна
    -0.13
    dash
    -0.13
    eview
    -0.13
    recommend
    -0.12
    à¸ķา
    -0.12
    iedad
    -0.12
    ondo
    -0.12
     ifndef
    -0.12
    choices
    -0.12
    POSITIVE LOGITS
     footing
    0.35
     niche
    0.28
     groove
    0.27
     balance
    0.26
     way
    0.25
     equilibrium
    0.23
     Balance
    0.22
     nich
    0.22
    ilibrium
    0.22
     voice
    0.21
    Activations Density 0.104%

    No Known Activations