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

    Neuronpedia

    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. 0-RES-JB
    11. 9432
    Prev
    Next
    INDEX
    Explanations

    phrases indicating ongoing events or situations

    oai_token-act-pair · gpt-3.5-turbo

    frequent mentions of the word "happening."

    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
     assess
    -0.64
    bern
    -0.61
    eters
    -0.57
    esses
    -0.57
    en
    -0.55
     inspect
    -0.55
    rel
    -0.55
     Des
    -0.54
     praises
    -0.54
     prepare
    -0.53
    POSITIVE LOGITS
     happening
    3.41
     occurring
    2.39
     happen
    2.01
     happened
    1.72
     happens
    1.61
     happ
    1.58
     unfolding
    1.47
     Happ
    1.36
     transpired
    1.25
     occur
    1.24
    Activations Density 0.013%

    No Known Activations