© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 1-LLAMASCOPE-RES-32K
    11. 26516
    Prev
    Next
    INDEX
    Explanations

    instances of the word "fail" in various forms and contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 1-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L1R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.1.hook_resid_post
    Hook Layer
    1
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    onto
    -0.16
    zdy
    -0.15
    ÙĤØ·
    -0.15
    žel
    -0.15
    \API
    -0.14
     *)((
    -0.14
    ismatch
    -0.14
     à¤ĩसस
    -0.14
    apat
    -0.14
    hel
    -0.13
    POSITIVE LOGITS
     Failed
    0.16
     HQ
    0.15
     Fail
    0.15
     rej
    0.15
     failed
    0.14
     Headquarters
    0.14
    Scenario
    0.14
    isma
    0.14
    omain
    0.14
    orsch
    0.14
    Activations Density 0.021%

    No Known Activations