© 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 EvalsExportsinterp-engineNEWAPI 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. 3-RES-JB
    11. 16729
    Prev
    Next
    INDEX
    Explanations

    words related to "Astrophysics" and similar terms

    oai_token-act-pair · gpt-3.5-turbo

    references to astronomy

    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
    ä¹ĭ
    -0.76
    cific
    -0.75
    flix
    -0.73
    peat
    -0.71
    iard
    -0.69
    backer
    -0.68
    inki
    -0.66
    ãģĭ
    -0.66
     Orig
    -0.64
     Assange
    -0.64
    POSITIVE LOGITS
    rament
    1.04
    roph
    0.99
    hma
    0.83
     Ast
    0.81
    acio
    0.80
    rop
    0.78
    igmat
    0.77
    Ast
    0.76
    ellation
    0.76
    ound
    0.76
    Activations Density 0.007%

    No Known Activations