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

    Neuronpedia

    Jacobian LensNEW
    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. 1-RES-JB
    11. 22045
    Prev
    Next
    INDEX
    Explanations

    phrases related to illegal activities or schemes

    oai_token-act-pair · gpt-3.5-turbo

    references to schemes or situations involving exploitation or fraud

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     turb
    -0.70
    borgh
    -0.68
     AFB
    -0.68
    amps
    -0.68
     sands
    -0.67
     castles
    -0.67
     acid
    -0.65
     quarters
    -0.65
    amo
    -0.64
     departments
    -0.62
    POSITIVE LOGITS
    eering
    1.88
    eers
    1.59
    eer
    1.46
     racket
    1.32
    acket
    0.94
    urtle
    0.80
    ablishment
    0.79
    nik
    0.78
    aminer
    0.76
    rious
    0.75
    Activations Density 0.015%

    No Known Activations