© 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. 1958
    Prev
    Next
    INDEX
    Explanations

    proper nouns, especially names of countries

    oai_token-act-pair · gpt-3.5-turbo

    words that end with specific suffixes indicating various grammatical forms and meanings

    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
     Pastebin
    -0.65
    Secure
    -0.63
     Bridgewater
    -0.60
     Waves
    -0.58
    staff
    -0.58
    ï¸ı
    -0.56
     Secure
    -0.56
    inus
    -0.56
    rador
    -0.55
     Eater
    -0.55
    POSITIVE LOGITS
    ROR
    0.84
    /-
    0.82
    verage
    0.80
    hesion
    0.74
    ricks
    0.69
    phia
    0.68
    ãĥĨ
    0.67
    ãĤ¨ãĥ«
    0.67
    vernight
    0.63
     representation
    0.61
    Activations Density 0.380%

    No Known Activations