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. 0-RES-JB
    6. 17305
    Prev
    Next
    INDEX
    Explanations

    phone numbers or dates written in the format MM/DD/YY

    oai_token-act-pair · gpt-3.5-turbo

    references to the number 88 and related numerical values

    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
     drag
    -0.89
     dragging
    -0.77
     Spart
    -0.75
     Gard
    -0.69
     Fra
    -0.68
     subsequ
    -0.68
    orter
    -0.68
     Parm
    -0.68
     Shepard
    -0.67
     dragged
    -0.67
    POSITIVE LOGITS
    88
    2.74
    89
    1.96
    888
    1.87
    889
    1.78
     88
    1.71
    87
    1.64
    885
    1.58
    882
    1.51
    1988
    1.49
    887
    1.49
    Activations Density 0.052%

    No Known Activations