© 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. 8-RES-JB
    11. 15095
    Prev
    Next
    INDEX
    Explanations

    instances of the word 'when' combined with actions or scenarios

    oai_token-act-pair · gpt-3.5-turbo

    conditional phrases that contain the word "when."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    went
    -0.69
    entry
    -0.68
    him
    -0.68
    eme
    -0.67
    yan
    -0.65
    hid
    -0.63
    guard
    -0.62
    rieve
    -0.62
    bi
    -0.61
    dump
    -0.61
    POSITIVE LOGITS
     compared
    1.27
     contrasted
    1.17
    soever
    1.09
     comparing
    1.03
     coupled
    1.01
     confronted
    0.96
     viewed
    0.95
     combined
    0.93
     considering
    0.92
     faced
    0.89
    Activations Density 0.155%

    No Known Activations