© 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. 26-LLAMASCOPE-RES-32K
    11. 19570
    Prev
    Next
    INDEX
    Explanations

    phrases that emphasize the concept of honesty or being truthful

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.26.hook_resid_post
    Hook Layer
    26
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    andas
    -0.16
    inz
    -0.15
    èĩº
    -0.15
    åı°
    -0.15
     addAction
    -0.14
    .writeln
    -0.14
    vyk
    -0.14
    pha
    -0.13
    atern
    -0.13
    Prompt
    -0.13
    POSITIVE LOGITS
     honest
    0.55
     Honest
    0.45
     frank
    0.42
     honesty
    0.41
     candid
    0.37
     perfectly
    0.32
     cand
    0.31
     Frank
    0.30
     honestly
    0.29
     truthful
    0.29
    Activations Density 0.030%

    No Known Activations