© 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. 4-RES-JB
    11. 24314
    Prev
    Next
    INDEX
    Explanations

    ordinal numbers combined with certain adjectives

    oai_token-act-pair · gpt-3.5-turbo

    ordinal numbers and rankings

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

    No Comments

    Negative Logits
    ENSE
    -0.63
    styles
    -0.62
    each
    -0.60
    md
    -0.60
    Sm
    -0.59
    Box
    -0.58
    agate
    -0.57
    don
    -0.57
    ourage
    -0.56
    åĪ
    -0.56
    POSITIVE LOGITS
    liest
    0.82
     casualty
    0.79
     conceivable
    0.78
     frontrunner
    0.75
     focal
    0.72
     beneficiary
    0.72
     centerpiece
    0.71
     busiest
    0.70
     antidote
    0.68
     culmination
    0.67
    Activations Density 0.221%

    No Known Activations