© 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. 22585
    Prev
    Next
    INDEX
    Explanations

    comparative phrases indicating a choice between two options

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "versus."

    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
    bris
    -0.64
    ŀ
    -0.62
     AUTH
    -0.62
     Freak
    -0.60
    FORM
    -0.57
     soon
    -0.53
    Ùħ
    -0.53
     Tatt
    -0.53
     Pieces
    -0.52
    full
    -0.52
    POSITIVE LOGITS
     versus
    3.75
     vs
    2.54
     Versus
    2.28
     Vs
    1.75
     compared
    1.43
    vs
    1.41
     VS
    1.19
     against
    1.07
     whereas
    1.04
    against
    0.98
    Activations Density 0.012%

    No Known Activations