© 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. 1-RES-JB
    11. 3229
    Prev
    Next
    INDEX
    Explanations

    words related to providing entertainment or amusement

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "entertain" and related forms, reflecting a focus on entertainment-related activities or concepts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     descent
    -0.73
     holiest
    -0.67
     prone
    -0.62
     broken
    -0.61
    installed
    -0.60
     Archangel
    -0.59
     uneven
    -0.59
     planted
    -0.59
     polymer
    -0.59
    graded
    -0.58
    POSITIVE LOGITS
     entertain
    1.37
    tainment
    1.21
     entertained
    1.14
    glers
    1.04
    ments
    1.03
    eers
    1.00
    eering
    0.98
    abilia
    0.98
     entert
    0.93
    atorium
    0.89
    Activations Density 0.006%

    No Known Activations