© 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. 2-RES-JB
    11. 5779
    Prev
    Next
    INDEX
    Explanations

    phrases related to a specific event called "Turn"

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "Turn" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    è¦ļéĨĴ
    -0.93
    mble
    -0.74
     franc
    -0.67
     borough
    -0.66
     Tibetan
    -0.65
    ãģª
    -0.63
     Lama
    -0.62
     dan
    -0.61
    ãĤĮ
    -0.60
    éĹ
    -0.59
    POSITIVE LOGITS
    turn
    0.93
    Turn
    0.90
    overs
    0.87
    bum
    0.78
     Turn
    0.76
    around
    0.75
    ners
    0.74
    Stop
    0.73
    stop
    0.72
     Turns
    0.72
    Activations Density 0.012%

    No Known Activations