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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 10-RES-JB
    11. 7937
    Prev
    Next
    INDEX
    Explanations

    words related to physical discomfort or trouble, such as "ghastly" or "rough"

    oai_token-act-pair · gpt-3.5-turbo

    the word "gh" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     clustered
    -0.67
     closest
    -0.64
    IAL
    -0.61
    kefeller
    -0.59
     minimized
    -0.59
     symp
    -0.59
     satell
    -0.58
     generalized
    -0.58
     similarly
    -0.58
     prescribed
    -0.57
    POSITIVE LOGITS
    ttp
    1.23
    ouls
    1.14
    arrison
    1.05
    orns
    1.03
    awk
    1.01
    tml
    1.00
    ouse
    0.97
    hhh
    0.95
    orne
    0.94
    olics
    0.94
    Activations Density 0.025%

    No Known Activations