© 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. 25-LLAMASCOPE-RES-32K
    11. 10482
    Prev
    Next
    INDEX
    Explanations

    instances of unique names, locations, and significant identifiers

    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
    âĢŀM
    -0.14
    âĢŀN
    -0.14
    ̣
    -0.14
     entirety
    -0.14
    @qq
    -0.13
    âĢŀA
    -0.13
    561
    -0.13
    âĢŀP
    -0.13
    ìł¸
    -0.13
    ÄĽÅ¾
    -0.12
    POSITIVE LOGITS
    .toolbox
    0.14
    antage
    0.14
    é»
    0.14
    atal
    0.13
     دع
    0.13
    usa
    0.13
    =$((
    0.13
    èĨľ
    0.12
    ickest
    0.12
    ivid
    0.12
    Activations Density 0.825%

    No Known Activations