© 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. 13114
    Prev
    Next
    INDEX
    Explanations

    phrases related to positive evaluation or endorsement

    oai_token-act-pair · gpt-3.5-turbo

    discussions or phrases emphasizing positive qualities or the word "good."

    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
    eters
    -0.85
    eds
    -0.79
    ople
    -0.77
    istan
    -0.72
    ratom
    -0.71
    aters
    -0.70
    idon
    -0.70
    NetMessage
    -0.69
    agos
    -0.69
    hip
    -0.68
    POSITIVE LOGITS
    enough
    1.31
     enough
    1.06
     chunk
    1.00
    sword
    0.98
     luck
    0.92
     ol
    0.90
     approximation
    0.87
     Samar
    0.87
     bye
    0.86
     idea
    0.84
    Activations Density 0.041%

    No Known Activations