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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 3-RES-JB
    6. 1358
    Prev
    Next
    INDEX
    Explanations

    terms related to specific regions or districts

    oai_token-act-pair · gpt-3.5-turbo

    references to regional organizations or entities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    iami
    -0.71
    livion
    -0.69
    olson
    -0.68
    ighty
    -0.66
    illary
    -0.64
    izabeth
    -0.63
    gerald
    -0.63
    âĶĢâĶĢ
    -0.62
     circulation
    -0.61
     vein
    -0.61
    POSITIVE LOGITS
     Holding
    0.98
     Abuse
    0.94
     Authority
    0.90
     Laboratories
    0.90
     Responsibility
    0.87
     Identification
    0.87
     Affect
    0.83
     Shooter
    0.83
    senal
    0.81
     Relations
    0.81
    Activations Density 0.082%

    No Known Activations