© 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. 23268
    Prev
    Next
    INDEX
    Explanations

    words related to events or occasions, particularly those related to debuts or first showings

    oai_token-act-pair · gpt-3.5-turbo

    references to premieres of shows, movies, or events

    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
    µ
    -0.70
    Beh
    -0.66
    etooth
    -0.60
    Boss
    -0.59
    let
    -0.58
    hab
    -0.58
     logic
    -0.57
     Beh
    -0.57
    bal
    -0.57
    ogged
    -0.57
    POSITIVE LOGITS
     premiere
    3.80
     premie
    2.48
     Premiere
    2.31
     premier
    1.88
     premiered
    1.79
     finale
    1.70
     debut
    1.50
     screenings
    1.18
     opener
    1.10
     airing
    1.10
    Activations Density 0.010%

    No Known Activations