© 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. 17610
    Prev
    Next
    INDEX
    Explanations

    references to storytelling elements and character development within a series

    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
    bij
    -0.16
    isay
    -0.15
    ìĭ±
    -0.14
    ividad
    -0.14
    å½ķ
    -0.14
    igli
    -0.14
    agnost
    -0.13
     FileMode
    -0.13
    ols
    -0.13
    éĮĦ
    -0.13
    POSITIVE LOGITS
     flesh
    0.30
     Flesh
    0.23
     introdu
    0.23
    Introduced
    0.22
     Establish
    0.21
     introduce
    0.20
     established
    0.20
     Introduced
    0.20
     establish
    0.20
     sho
    0.20
    Activations Density 0.127%

    No Known Activations