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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 30-LLAMASCOPE-RES-32K
    11. 23696
    Prev
    Next
    INDEX
    Explanations

    expressions related to moral evaluations and legality

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.30.hook_resid_post
    Hook Layer
    30
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    given
    -0.18
    Ñħи
    -0.18
    istrovstvÃŃ
    -0.17
     Given
    -0.16
     given
    -0.16
     Considering
    -0.16
     ><?
    -0.16
    eldom
    -0.16
    VF
    -0.15
     considering
    -0.15
    POSITIVE LOGITS
     because
    0.38
     BE
    0.36
     precisely
    0.33
    because
    0.31
     Because
    0.31
     porque
    0.31
    prec
    0.31
    Because
    0.30
     поÑĤомÑĥ
    0.29
     Prec
    0.28
    Activations Density 0.147%

    No Known Activations