© 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. 3321
    Prev
    Next
    INDEX
    Explanations

    highly positive or superlative expressions

    oai_token-act-pair · gpt-3.5-turbo

    expressions of personal experience or opinion

    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
     Bronze
    -0.60
    clerosis
    -0.59
     Schiff
    -0.55
     loophole
    -0.55
     markup
    -0.54
     bidder
    -0.54
     consolidation
    -0.53
     POL
    -0.52
     mixes
    -0.52
     Stub
    -0.52
    POSITIVE LOGITS
    20439
    0.95
    ĨĴ
    0.85
    ģĸ
    0.83
    ãĥ¼ãĥĨ
    0.80
    ãĤ´
    0.77
    erest
    0.75
    selves
    0.74
    ovie
    0.73
    ª
    0.72
    given
    0.72
    Activations Density 0.152%

    No Known Activations