© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 22307
    Prev
    Next
    INDEX
    Explanations

    words related to advice or recommendations for actions

    oai_token-act-pair · gpt-3.5-turbo

    phrases that indicate recommendations or suggestions for specific actions

    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
    lance
    -0.84
    ãĥ¼ãĥ³
    -0.76
    ãĤ¼ãĤ¦ãĤ¹
    -0.76
    ivation
    -0.74
    00007
    -0.69
    ppa
    -0.69
    aucus
    -0.68
    NAS
    -0.67
     istg
    -0.67
    nea
    -0.66
    POSITIVE LOGITS
     varying
    0.99
     various
    0.93
     enhance
    0.81
     improve
    0.81
     mitigate
    0.78
     conceal
    0.77
     differing
    0.77
     mathemat
    0.77
     different
    0.77
     strengthen
    0.74
    Activations Density 0.445%

    No Known Activations