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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 0-RES-JB
    11. 22273
    Prev
    Next
    INDEX
    Explanations

    titles or names containing the sequence "Ke"

    oai_token-act-pair · gpt-3.5-turbo

    specific names or identifiers associated with individuals or entities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     combat
    -0.66
     online
    -0.66
     Norm
    -0.66
     sup
    -0.64
     conduit
    -0.64
     grooming
    -0.63
     Horowitz
    -0.62
     placement
    -0.62
     PA
    -0.62
     breakdown
    -0.61
    POSITIVE LOGITS
    Ke
    3.82
    Ka
    1.50
    Ze
    1.20
    Ge
    1.14
    Haw
    1.10
    Dar
    1.08
    Hay
    1.08
    Ti
    1.08
     Kejriwal
    1.07
    Bay
    1.07
    Activations Density 0.015%

    No Known Activations