© 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. 4-RES-JB
    11. 14931
    Prev
    Next
    INDEX
    Explanations

    names or words with the letter sequence "j" followed by a number, such as "j9" or "j10"

    oai_token-act-pair · gpt-3.5-turbo

    the letter "j"

    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
     Lauder
    -0.79
     outgoing
    -0.68
     bleach
    -0.67
     dispers
    -0.66
    BALL
    -0.64
     Scotia
    -0.64
     productive
    -0.63
     Samoa
    -0.63
     dispersed
    -0.63
     prolific
    -0.63
    POSITIVE LOGITS
    ournals
    1.48
    acket
    1.34
    itsu
    1.30
    ealous
    1.28
    unction
    1.25
    ordan
    1.18
    unal
    1.14
    utsu
    1.10
    ungle
    1.09
    ihad
    1.08
    Activations Density 0.030%

    No Known Activations