© 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. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 51856
    Prev
    Next
    INDEX
    Explanations

    Amsterdam

    np_max-act · gemini-2.0-flash

    language that signals procedural steps and temporal transitions in a process, such as initiating, sequencing, or moving between stages in instructions or narratives.

    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_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    力
    -0.07
    26
    -0.07
    _LOOKUP
    -0.07
    เตร
    -0.06
     Из
    -0.06
     مناطق
    -0.06
    -envelope
    -0.06
     Fresno
    -0.06
     واحد
    -0.06
     Valle
    -0.06
    POSITIVE LOGITS
     bearer
    0.07
    .contacts
    0.07
    (level
    0.06
     digestion
    0.06
     spraying
    0.06
     абсолют
    0.06
    lemn
    0.06
     OAuth
    0.06
     пол
    0.06
     Publish
    0.06
    Activations Density 0.649%

    No Known Activations