Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 7-RESID-POST-AA
    6. 75008
    Prev
    Next
    INDEX
    Explanations

    Punctuation

    np_max-act · gemini-2.0-flash

    end-of-sentence or clause boundary tokens—words with attached punctuation (commas, periods, quotes/parentheses) and special end-of-turn markers.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_7/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.7.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    order
    -0.07
     strangers
    -0.07
    atak
    -0.07
     取
    -0.06
    Calc
    -0.06
    wagon
    -0.06
    classify
    -0.06
    {{--
    -0.06
    38
    -0.06
     wealthy
    -0.06
    POSITIVE LOGITS
    Ctrls
    0.07
     tha
    0.06
     Marseille
    0.06
     профессиональ
    0.06
     iken
    0.06
    _MEMBER
    0.06
     Afghan
    0.06
    .tagName
    0.06
     mocks
    0.06
     edilmiş
    0.06
    Activations Density 0.711%

    No Known Activations