© 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. 16857
    Prev
    Next
    INDEX
    Explanations

    words related to darkness and dimly lit environments

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "dark."

    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
    payers
    -0.77
    hips
    -0.70
     reimbursement
    -0.70
    onse
    -0.70
    olo
    -0.67
    ivari
    -0.67
    itis
    -0.66
    "]=>
    -0.65
     IEEE
    -0.65
    backer
    -0.65
    POSITIVE LOGITS
     dark
    3.51
    dark
    2.60
     darker
    2.52
     darkest
    2.31
     darkness
    2.22
    Dark
    2.18
     Dark
    2.17
     DARK
    2.03
     darkened
    1.87
     bright
    1.60
    Activations Density 0.013%

    No Known Activations