© 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. 25-LLAMASCOPE-RES-32K
    11. 13943
    Prev
    Next
    INDEX
    Explanations

    negative statements about feeling isolated or unsupported

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.25.hook_resid_post
    Hook Layer
    25
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ÏĢοÏĤ
    -0.14
    MOTE
    -0.14
    ê°ľë¥¼
    -0.14
    arken
    -0.14
    ogh
    -0.14
    IGHLIGHT
    -0.14
    ajÃŃ
    -0.14
    licit
    -0.13
    ilia
    -0.13
    ắp
    -0.13
    POSITIVE LOGITS
     alone
    0.38
     AL
    0.33
    Al
    0.31
     Alone
    0.29
    alone
    0.28
    _al
    0.27
     along
    0.27
     sol
    0.26
    AL
    0.26
    -Al
    0.26
    Activations Density 0.030%

    No Known Activations