© 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. 41665
    Prev
    Next
    INDEX
    Explanations

    medical cases

    np_max-act · gemini-2.0-flash

    The neuron activates on numbers, comparison signs, and outcome‐related terms in clinical reports—i.e. when the text gives patient counts, percentages, measurements, or treatment results.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    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
     Comics
    -0.07
    긔
    -0.07
    Ben
    -0.06
    Pale
    -0.06
    ंद
    -0.06
    ition
    -0.06
    öst
    -0.06
     pill
    -0.06
     Rad
    -0.06
     PAL
    -0.06
    POSITIVE LOGITS
    particles
    0.07
     jsonObject
    0.07
     d
    0.07
    toArray
    0.07
     strtoupper
    0.07
     });
↵
↵
    0.07
    .land
    0.06
    	U
    0.06
    ])[
    0.06
    (py
    0.06
    Activations Density 0.063%

    No Known Activations