© 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
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 89690
    Prev
    Next
    INDEX
    Explanations

    Tangible/Abstract

    np_max-act · gemini-2.0-flash

    The neuron specializes in spotting mentions of abstract or conceptual definitions—especially the word “abstract” and its context in class-definition lines.

    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
     Substitute
    -0.06
     interle
    -0.06
     TRY
    -0.06
     ASE
    -0.06
     divergence
    -0.06
     supervised
    -0.06
     slamming
    -0.05
     Blocking
    -0.05
     Belly
    -0.05
    Merge
    -0.05
    POSITIVE LOGITS
     clandest
    0.07
     conspicuous
    0.07
    .getClassName
    0.06
     فنی
    0.06
    .getPassword
    0.06
     карт
    0.06
     genres
    0.06
    _UNSIGNED
    0.06
     destroyer
    0.06
    .ret
    0.06
    Activations Density 0.029%

    No Known Activations