© 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. 10-RES-JB
    11. 5029
    Prev
    Next
    INDEX
    Explanations

    comparative phrases (e.g., "higher than," "not very high")

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating high rates or frequencies

    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
    otti
    -0.78
     dishes
    -0.71
    attach
    -0.69
     motions
    -0.68
     Franch
    -0.66
    ande
    -0.66
    odor
    -0.64
     Craigslist
    -0.62
    ologue
    -0.61
    akeru
    -0.60
    POSITIVE LOGITS
     higher
    1.36
    higher
    1.33
    Higher
    1.31
     negligible
    1.29
     low
    1.24
     lower
    1.21
     inflated
    1.20
     underestimated
    1.18
     overest
    1.17
     proportional
    1.15
    Activations Density 0.348%

    No Known Activations