© 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. 17-LLAMASCOPE-RES-131K
    11. 92105
    Prev
    Next
    INDEX
    Explanations

    conjunctions, particularly "and," along with references to time and money

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    itty
    -0.07
    geme
    -0.07
    unct
    -0.07
    elopment
    -0.07
     WaitForSeconds
    -0.07
    aben
    -0.07
     tarde
    -0.06
    olet
    -0.06
    owan
    -0.06
    otto
    -0.06
    POSITIVE LOGITS
     effort
    0.10
     resources
    0.08
     money
    0.07
     dollars
    0.07
     Vladim
    0.06
     space
    0.06
     Ingram
    0.06
     time
    0.06
     prob
    0.06
    agan
    0.06
    Activations Density 0.011%

    No Known Activations