© 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. 18574
    Prev
    Next
    INDEX
    Explanations

    the phrase "To be" followed by an opinion or explanation

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "to be."

    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
     Prescott
    -0.60
    rador
    -0.59
     Stain
    -0.56
    ria
    -0.56
    iasco
    -0.56
     nap
    -0.55
     residues
    -0.55
     Hep
    -0.55
     layers
    -0.54
     residue
    -0.53
    POSITIVE LOGITS
     able
    0.95
     honest
    0.95
     sure
    0.88
     blunt
    0.83
     frank
    0.83
     forgiven
    0.79
     eligible
    0.77
    heading
    0.75
     reminded
    0.74
    ardless
    0.73
    Activations Density 0.053%

    No Known Activations