© 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. 0-RES-JB
    11. 16556
    Prev
    Next
    INDEX
    Explanations

    references to cylindrical objects that can be rolled

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "rolls" in various contexts

    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
    heter
    -0.90
    Streamer
    -0.65
    Parameter
    -0.64
    angan
    -0.61
    asm
    -0.60
     Hiroshima
    -0.60
    athom
    -0.59
    atable
    -0.59
    ĨĴ
    -0.58
    ext
    -0.58
    POSITIVE LOGITS
     rolls
    4.06
     Rolls
    2.18
     roll
    2.17
     rolled
    1.94
    roll
    1.72
     Roll
    1.61
     rolling
    1.51
    Roll
    1.40
     scrolls
    1.23
     folds
    1.22
    Activations Density 0.009%

    No Known Activations