© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 2-RES-JB
    11. 7193
    Prev
    Next
    INDEX
    Explanations

    proper nouns or terms related to law and legal cases

    oai_token-act-pair · gpt-3.5-turbo

    references to individuals named Abram and topics related to antitrust issues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    OPLE
    -0.78
     Dinosaur
    -0.74
     versa
    -0.66
     Sikh
    -0.66
     Goose
    -0.65
     Towns
    -0.63
    lees
    -0.63
     Fukushima
    -0.61
    iness
    -0.61
    iking
    -0.60
    POSITIVE LOGITS
     Abram
    0.89
    mentation
    0.85
    Īè
    0.85
    AX
    0.83
    aic
    0.81
    furt
    0.80
    riad
    0.79
    Ĥİ
    0.78
    eering
    0.77
    ovich
    0.76
    Activations Density 0.032%

    No Known Activations