© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 15-LLAMASCOPE-RES-32K
    11. 13677
    Prev
    Next
    INDEX
    Explanations

    instances of the word "pretend" and its variations to identify themes of deception or acting

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    the verb "pretend" or "pretending" along with its immediate context (often followed by "to" or "not to").

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 15-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L15R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    /fw
    -0.17
    _INTR
    -0.17
    ackers
    -0.15
    IEW
    -0.15
    heid
    -0.15
    adx
    -0.14
    átu
    -0.14
    dain
    -0.14
    odge
    -0.14
    oggler
    -0.13
    POSITIVE LOGITS
    relative
    0.15
     Hayden
    0.15
     Clamp
    0.15
    azzi
    0.15
     mistake
    0.14
     relative
    0.14
    Relative
    0.14
    bah
    0.14
    unity
    0.14
    ô
    0.14
    Activations Density 0.033%

    No Known Activations