© 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. 1065
    Prev
    Next
    INDEX
    Explanations

    names or terms related to linguistics

    oai_token-act-pair · gpt-3.5-turbo

    text or phrases related to identifying information, particularly names and specific terminology in discussions about complex subjects

    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
     Coleman
    -0.83
     Eden
    -0.74
     Mous
    -0.71
     Archangel
    -0.68
    Cole
    -0.68
     Rahman
    -0.66
     Ingram
    -0.66
     Mitchell
    -0.66
     Liang
    -0.65
     Lanka
    -0.65
    POSITIVE LOGITS
    sf
    2.90
    ingu
    1.84
    sv
    1.53
    sb
    1.40
    sd
    1.34
    sr
    1.22
     Mayer
    1.16
    erd
    1.11
    ansk
    1.11
     unch
    1.09
    Activations Density 0.063%

    No Known Activations