Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 8-RES-JB
    6. 2986
    Prev
    Next
    INDEX
    Explanations

    mentions of people receiving beneficial outcomes or opportunities

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating the presence of the article "a" and assess its frequency

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    phrases containing any conjugation of "get" followed by the wor "a" (e.g. "gotten a", "get a")

    Explanation Uploaded by User
    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
    agree
    -0.77
    anism
    -0.68
    Events
    -0.68
    Ec
    -0.67
    always
    -0.67
    Types
    -0.66
    Iss
    -0.65
     endeavors
    -0.65
    ources
    -0.63
    errors
    -0.63
    POSITIVE LOGITS
     glimpse
    1.19
     foothold
    1.19
    hold
    1.12
     lot
    1.08
     chance
    1.01
     bunch
    0.95
     haircut
    0.90
     handful
    0.86
     clue
    0.85
     couple
    0.84
    Activations Density 0.198%

    No Known Activations