© 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. 13967
    Prev
    Next
    INDEX
    Explanations

    phrases expressing criticism or frustration towards decisions made by sports teams or individuals

    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
    201
    -0.24
     twitter
    -0.21
     Drone
    -0.20
    Û²Û°Û±
    -0.20
     drone
    -0.20
     graphene
    -0.19
     Occupy
    -0.19
     Zika
    -0.19
     tweeting
    -0.18
     Obama
    -0.18
    POSITIVE LOGITS
    199
    0.74
    Û±Û¹Û¹
    0.51
     Clinton
    0.43
    Clinton
    0.35
     Clintons
    0.29
     fax
    0.27
     millennium
    0.27
     AOL
    0.25
     AIDS
    0.23
     Millennium
    0.23
    Activations Density 0.669%

    No Known Activations