© 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. 0-RES-JB
    11. 4026
    Prev
    Next
    INDEX
    Explanations

    markup elements like underscored and categorized data

    oai_token-act-pair · gpt-3.5-turbo

    categories or headings in structured data

    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
     Temper
    -0.80
     Fiorina
    -0.77
    hov
    -0.74
     Lowry
    -0.72
     Debor
    -0.68
    Depth
    -0.68
     Neh
    -0.67
     Kepler
    -0.65
     Manip
    -0.64
     Muss
    -0.64
    POSITIVE LOGITS
     favourites
    0.72
    arters
    0.70
    _(
    0.68
    ookie
    0.65
    assic
    0.64
     whilst
    0.63
    aiden
    0.63
     apologised
    0.63
     honoured
    0.63
    fortable
    0.62
    Activations Density 0.000%

    No Known Activations