© 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. 295
    Prev
    Next
    INDEX
    Explanations

    references to conclusions or summaries

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "bottom line."

    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
    sure
    -0.79
    agine
    -0.72
    kson
    -0.72
    Äĩ
    -0.71
    thouse
    -0.69
    encer
    -0.68
    areth
    -0.67
    cia
    -0.66
    atche
    -0.66
    issance
    -0.64
    POSITIVE LOGITS
     drawer
    0.99
    most
    0.91
     shelf
    0.86
     bracket
    0.84
    chart
    0.83
     margin
    0.80
    less
    0.79
    lander
    0.77
    hand
    0.76
    lined
    0.76
    Activations Density 0.022%

    No Known Activations