© 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. 16-LLAMASCOPE-RES-32K
    11. 20688
    Prev
    Next
    INDEX
    Explanations

    expressions related to psychological or emotional concepts, particularly those involving affirmation and conflict

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    egin
    -0.16
    uzz
    -0.15
    éĤ¦
    -0.14
    jang
    -0.14
    nge
    -0.13
     bì
    -0.13
    aktu
    -0.13
    iko
    -0.13
    tw
    -0.13
    há
    -0.13
    POSITIVE LOGITS
     besides
    0.24
     such
    0.21
     including
    0.21
     aside
    0.21
    :
    0.20
    :↵
    0.20
     include
    0.18
     ranging
    0.18
    :↵↵
    0.18
    bes
    0.17
    Activations Density 0.412%

    No Known Activations