© 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. 0-RES-JB
    11. 16834
    Prev
    Next
    INDEX
    Explanations

    thinking, dedication, and loyalty-related terms or phrases

    oai_token-act-pair · gpt-3.5-turbo

    words related to eligibility and service

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     altar
    -0.85
     Halloween
    -0.79
     Masquerade
    -0.78
    Hunt
    -0.73
     tar
    -0.73
    tips
    -0.73
     Olsen
    -0.69
    WT
    -0.68
     Thanksgiving
    -0.68
     herpes
    -0.68
    POSITIVE LOGITS
    ellig
    3.59
    serv
    2.22
     serv
    1.63
    Serv
    1.38
     intellig
    1.36
     Serv
    1.30
    elligent
    1.30
    glas
    1.29
    redible
    1.23
    vernment
    1.22
    Activations Density 0.041%

    No Known Activations