© 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. 20843
    Prev
    Next
    INDEX
    Explanations

    references to specific terms like "FC" with varying degrees of emphasis

    oai_token-act-pair · gpt-3.5-turbo

    references to facilitated communication (FC) and related terminology

    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
    Downloadha
    -0.76
     Dru
    -0.71
     Ples
    -0.65
     heck
    -0.64
    lihood
    -0.64
     Schne
    -0.63
     Gry
    -0.61
     Carmen
    -0.61
     deaf
    -0.61
     speedy
    -0.60
    POSITIVE LOGITS
    urrent
    0.95
    ombat
    0.92
    riad
    0.88
    ranch
    0.87
    FC
    0.87
    asting
    0.87
    ruits
    0.86
    ridges
    0.84
    erning
    0.83
    ombo
    0.83
    Activations Density 0.006%

    No Known Activations