© 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. 7982
    Prev
    Next
    INDEX
    Explanations

    phrases related to the action of writing or creating documents or plans

    oai_token-act-pair · gpt-3.5-turbo

    instances of drafting and crafting activities

    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
    AppData
    -0.78
    unks
    -0.69
     Virgin
    -0.67
    osterone
    -0.66
     emitted
    -0.66
    ocks
    -0.65
    nant
    -0.64
    amba
    -0.64
     resident
    -0.63
     TMZ
    -0.63
    POSITIVE LOGITS
     drafting
    3.31
     crafting
    2.23
     designing
    1.87
     selecting
    1.43
     framing
    1.35
     assembling
    1.35
     drafts
    1.29
     arranging
    1.27
     negotiating
    1.25
     constructing
    1.23
    Activations Density 0.029%

    No Known Activations