© 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
    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. 8216
    Prev
    Next
    INDEX
    Explanations

    names of individuals

    oai_token-act-pair · gpt-3.5-turbo

    proper names of individuals, likely focusing on notable figures or characters

    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
    Sax
    -0.65
     hindsight
    -0.65
    yip
    -0.63
    ements
    -0.63
    GoldMagikarp
    -0.63
     conclud
    -0.61
     foothold
    -0.61
    idental
    -0.61
     guts
    -0.60
    ardless
    -0.60
    POSITIVE LOGITS
     herself
    0.89
     nursing
    0.76
    otte
    0.76
    hyde
    0.72
    Jane
    0.72
     nurse
    0.71
     Duchess
    0.71
    lene
    0.70
    deen
    0.69
     maid
    0.68
    Activations Density 0.610%

    No Known Activations