© 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. 10-RES-JB
    11. 18706
    Prev
    Next
    INDEX
    Explanations

    people's names starting with the letter "H" and "M"

    oai_token-act-pair · gpt-3.5-turbo

    initials and names of notable individuals, particularly actors

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     unic
    -0.72
     Romanian
    -0.67
    âĶĢâĶĢ
    -0.62
     Freddie
    -0.62
     Geh
    -0.60
     Sahara
    -0.58
    RIC
    -0.56
     Mae
    -0.56
     orphans
    -0.56
    cest
    -0.55
    POSITIVE LOGITS
    Ĥ¬
    0.83
    ĵĺ
    0.75
    otos
    0.74
    ourke
    0.71
    ramer
    0.71
    antz
    0.70
    bett
    0.69
    lake
    0.69
    lamm
    0.68
    zinski
    0.68
    Activations Density 0.104%

    No Known Activations