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. 11823
    Prev
    Next
    INDEX
    Explanations

    numerical values representing the quantity of deaths

    oai_token-act-pair · gpt-3.5-turbo

    references to mortality statistics

    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
    arity
    -0.76
    Alpha
    -0.76
     Plex
    -0.71
    PLA
    -0.71
    Cola
    -0.70
     Libre
    -0.69
    Stud
    -0.67
    OPER
    -0.62
    TAG
    -0.62
     Cube
    -0.62
    POSITIVE LOGITS
     deaths
    0.89
     toll
    0.88
    hooting
    0.84
     overdoses
    0.79
    pine
    0.77
    afe
    0.76
    umatic
    0.75
    ffen
    0.74
    icides
    0.74
     inquest
    0.73
    Activations Density 0.031%

    No Known Activations