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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 10-RES-JB
    11. 21801
    Prev
    Next
    INDEX
    Explanations

    information about applying for programs or services

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to learning and applying for programs or information

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Freak
    -0.78
    Redditor
    -0.71
     dstg
    -0.69
     Buk
    -0.69
    Untitled
    -0.67
    HAHA
    -0.66
     overpowered
    -0.66
    fuck
    -0.66
    luaj
    -0.65
     [+
    -0.65
    POSITIVE LOGITS
     broch
    1.12
     Tickets
    1.03
     details
    0.98
     information
    0.97
     schedule
    0.96
     Schedule
    0.95
     schedules
    0.94
     Directions
    0.93
     Locations
    0.93
     availability
    0.92
    Activations Density 0.263%

    No Known Activations