© 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. 25-LLAMASCOPE-RES-32K
    11. 20489
    Prev
    Next
    INDEX
    Explanations

    terms related to the interaction and effects of drugs, particularly in the context of safety and abuse potential

    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
     spontaneously
    -0.15
    lect
    -0.14
    aterno
    -0.14
     compensated
    -0.14
    etr
    -0.14
    HX
    -0.13
     spontaneous
    -0.13
     {$
    -0.13
    arden
    -0.13
    adt
    -0.13
    POSITIVE LOGITS
     eup
    0.27
     psycho
    0.23
     sed
    0.23
     recreational
    0.21
     effects
    0.21
    effects
    0.20
     Buzz
    0.20
     abused
    0.20
     Effects
    0.19
    Effects
    0.19
    Activations Density 0.073%

    No Known Activations