© 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. 16787
    Prev
    Next
    INDEX
    Explanations

    contractions where the word "not" is followed by the apostrophe t

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "don't" or its variations

    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
     production
    -0.58
     populated
    -0.57
     along
    -0.55
     guided
    -0.54
     management
    -0.52
     lens
    -0.51
     accounting
    -0.49
     cur
    -0.49
     manned
    -0.49
     rose
    -0.49
    POSITIVE LOGITS
    't
    3.50
    NOT
    1.17
    \'
    1.16
    ´
    1.04
    ned
    0.99
    ',"
    0.98
    na
    0.93
    ÃĥÃĤ
    0.90
    doesn
    0.90
    etsk
    0.89
    Activations Density 0.097%

    No Known Activations