© 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
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 3-RES-JB
    6. 15422
    Prev
    Next
    INDEX
    Explanations

    proper nouns related to individuals, such as names

    oai_token-act-pair · gpt-3.5-turbo

    references to specific individuals, particularly those named "Pryor," "Benson," and "Patel."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Columb
    -0.78
    fare
    -0.77
    lehem
    -0.72
    igm
    -0.71
    achusetts
    -0.67
    amen
    -0.66
    istance
    -0.64
    ablishment
    -0.64
    ²
    -0.63
    icas
    -0.63
    POSITIVE LOGITS
     Pryor
    0.91
    hurst
    0.79
    folios
    0.72
    ridge
    0.71
    kus
    0.65
    hawk
    0.65
    Reloaded
    0.64
    backer
    0.64
    archment
    0.62
     dictate
    0.61
    Activations Density 0.017%

    No Known Activations