© 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. 7-RES-JB
    11. 17802
    Prev
    Next
    INDEX
    Explanations

    special characters and formatting patterns, potentially indicating that the neuron is looking for specific contexts or code structures rather than text content

    oai_token-act-pair · gpt-3.5-turbo

    content related to web links and URLs

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    7
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     inev
    -0.82
     spoil
    -0.77
     baggage
    -0.74
     diseng
    -0.74
     fatig
    -0.73
     redress
    -0.71
     invol
    -0.70
     charm
    -0.70
     habit
    -0.69
     dreaded
    -0.68
    POSITIVE LOGITS
    https
    1.08
    http
    1.02
    姫
    0.96
     Accessed
    0.94
    facebook
    0.94
    DragonMagazine
    0.93
    Example
    0.93
    Website
    0.91
    ================================================================
    0.90
    forums
    0.89
    Activations Density 0.135%

    No Known Activations