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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 3-RES-JB
    11. 6023
    Prev
    Next
    INDEX
    Explanations

    phrases related to surprising or strange occurrences

    oai_token-act-pair · gpt-3.5-turbo

    elements of surprise or irony in situations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    verning
    -0.80
    ederation
    -0.76
    esses
    -0.71
     equivalents
    -0.69
    acion
    -0.68
     subsidiaries
    -0.68
    assadors
    -0.67
     freely
    -0.65
    IDA
    -0.65
     habitable
    -0.64
    POSITIVE LOGITS
     Clown
    0.72
    etheless
    0.67
     miscarriage
    0.66
     omission
    0.60
    kid
    0.60
     Sturgeon
    0.60
     scratches
    0.59
     Solitaire
    0.59
    ovie
    0.58
     miscar
    0.58
    Activations Density 1.229%

    No Known Activations