© 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. 4-RES-JB
    11. 9032
    Prev
    Next
    INDEX
    Explanations

    descriptions of various statistics and opinions

    oai_token-act-pair · gpt-3.5-turbo

    statistical data and public opinion regarding political affiliations

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

    No Comments

    Negative Logits
     pus
    -0.62
    çīĪ
    -0.62
     erection
    -0.57
    rave
    -0.57
    æĦ
    -0.57
     airst
    -0.55
     fulfil
    -0.55
     layouts
    -0.54
    hai
    -0.54
     rend
    -0.53
    POSITIVE LOGITS
     Gallup
    0.92
     Percentage
    0.88
     averaging
    0.87
     Average
    0.85
    Average
    0.85
     statistically
    0.85
    verages
    0.85
     Median
    0.84
    adjusted
    0.82
     median
    0.81
    Activations Density 1.901%

    No Known Activations