© 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. 31-LLAMASCOPE-RES-32K
    11. 18500
    Prev
    Next
    INDEX
    Explanations

    specific patterns or occurrences of the character "Ñ" in various forms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ida
    -0.18
    genic
    -0.16
    cial
    -0.16
    ipp
    -0.15
    alin
    -0.15
    è²
    -0.14
     Kick
    -0.14
    ern
    -0.14
    ivas
    -0.14
    gh
    -0.14
    POSITIVE LOGITS
    нÑĨиклопед
    0.29
    леменÑĤ
    0.26
    л
    0.26
    вол
    0.23
    кÑģп
    0.21
    неÑĢг
    0.21
    коном
    0.19
    н
    0.19
    неÑĢ
    0.18
    lek
    0.18
    Activations Density 0.003%

    No Known Activations