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. 17590
    Prev
    Next
    INDEX
    Explanations

    mentions of a specific person, possibly a public figure or celebrity

    oai_token-act-pair · gpt-3.5-turbo

    references to the concept of "super PACs" and influential individuals associated with them

    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
     Bound
    -0.66
    vez
    -0.65
     MPG
    -0.64
     euth
    -0.64
     unlaw
    -0.62
    urb
    -0.60
     reversible
    -0.60
     Mate
    -0.60
     suicide
    -0.60
     Mort
    -0.59
    POSITIVE LOGITS
    ãģĭ
    2.13
    ãģĮ
    1.61
    hak
    1.35
    PAC
    1.21
    ulously
    1.07
    cz
    1.02
    ãģķ
    0.99
     painfully
    0.96
    arted
    0.88
    vous
    0.87
    Activations Density 0.023%

    No Known Activations