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

    Neuronpedia

    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. 3-RES-JB
    11. 9038
    Prev
    Next
    INDEX
    Explanations

    the word "clamoring" in different forms

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "clam" and related word forms

    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
    selection
    -0.70
    Georgia
    -0.68
     Agency
    -0.67
    ĻĤ
    -0.66
    stood
    -0.65
     Advisor
    -0.64
    hani
    -0.64
    gypt
    -0.62
    »Ĵ
    -0.61
    basketball
    -0.60
    POSITIVE LOGITS
     clam
    1.19
    borgh
    0.88
    pton
    0.85
    mong
    0.83
    bug
    0.81
    eless
    0.80
    orious
    0.79
    fish
    0.79
    iously
    0.76
    usters
    0.75
    Activations Density 0.005%

    No Known Activations