© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 7557
    Prev
    Next
    INDEX
    Explanations

    references to clerks or individuals working as clerks in various contexts

    oai_token-act-pair · gpt-3.5-turbo

    references to legal clerks or their roles

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    NAT
    -0.70
     Aman
    -0.69
     Imper
    -0.68
     Zero
    -0.67
    positive
    -0.65
    ures
    -0.64
    asts
    -0.62
    Zero
    -0.62
    URES
    -0.61
     DAM
    -0.61
    POSITIVE LOGITS
     clerk
    3.73
     clerks
    3.23
     Clerk
    2.87
     cler
    1.45
     solicitor
    1.39
     courthouse
    1.26
     bartender
    1.24
     attendant
    1.23
     treasurer
    1.22
     Registrar
    1.17
    Activations Density 0.037%

    No Known Activations