© 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. 4-RES-JB
    11. 15393
    Prev
    Next
    INDEX
    Explanations

    proper names, particularly the name "Levi"

    oai_token-act-pair · gpt-3.5-turbo

    mentions of specific names and locations

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

    No Comments

    Negative Logits
    ername
    -0.89
    itives
    -0.81
    lege
    -0.81
    enburg
    -0.79
    rish
    -0.78
    unin
    -0.76
    nam
    -0.74
    nesses
    -0.74
    staff
    -0.73
    law
    -0.72
    POSITIVE LOGITS
    andro
    1.12
    phia
    0.78
     mosqu
    0.74
    angelo
    0.74
    abetic
    0.69
    ynthesis
    0.68
    earchers
    0.67
     Shard
    0.67
     Tsukuyomi
    0.67
    ++++
    0.66
    Activations Density 0.026%

    No Known Activations