© 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. 8-RES-JB
    11. 6487
    Prev
    Next
    INDEX
    Explanations

    numbers and quantities described in a context of events or situations

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "another" and similar terms indicating additions or repetitions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ouls
    -0.75
    bows
    -0.75
    etc
    -0.67
    ipedia
    -0.67
    \-
    -0.63
     budgets
    -0.62
    Wiki
    -0.62
    icons
    -0.61
    çļ
    -0.61
     Wenger
    -0.61
    POSITIVE LOGITS
    worldly
    1.11
     batch
    1.04
     unnamed
    1.03
     unidentified
    0.92
     undisclosed
    0.90
     pair
    0.86
     dozen
    0.82
     trio
    0.79
     handful
    0.79
     group
    0.76
    Activations Density 0.193%

    No Known Activations