© 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. 0-RES-JB
    11. 15665
    Prev
    Next
    INDEX
    Explanations

    mentions of various colleges and institutions named "Hillsdale."

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "dale."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     partic
    -0.80
     Frag
    -0.72
     fingerprints
    -0.69
     prot
    -0.66
     favoring
    -0.66
     iso
    -0.63
     cooperating
    -0.62
     intensive
    -0.62
     strict
    -0.61
     Fu
    -0.61
    POSITIVE LOGITS
    dale
    4.62
    vale
    1.32
    dal
    1.29
    ville
    1.22
    brook
    1.19
    endale
    1.14
    combe
    1.12
    leigh
    1.10
    hurst
    1.06
    idge
    1.04
    Activations Density 0.015%

    No Known Activations