© 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. 4-RES-JB
    11. 2813
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "Ace" along with a high activation value indicating a strong association

    oai_token-act-pair · gpt-3.5-turbo

    references to the term "Ace" in various contexts

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

    No Comments

    Negative Logits
    isations
    -0.72
    eners
    -0.71
    icity
    -0.70
    FORMATION
    -0.70
    usalem
    -0.69
    arily
    -0.68
    ained
    -0.68
    attribute
    -0.68
    achu
    -0.68
    umers
    -0.68
    POSITIVE LOGITS
     Ace
    0.97
     Hardware
    0.81
    illary
    0.77
     Ventura
    0.76
    cean
    0.75
    idon
    0.74
    bay
    0.74
    Ĥª
    0.70
    reon
    0.69
    alys
    0.68
    Activations Density 0.005%

    No Known Activations