© 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. 2055
    Prev
    Next
    INDEX
    Explanations

    "negative actions or criticisms related to people or entities."

    oai_token-act-pair · gpt-3.5-turbo

    actions or events that involve criticism or controversy

    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
    Jews
    -0.70
    iphate
    -0.68
    assic
    -0.67
    itamin
    -0.61
    ãĥ
    -0.59
    idae
    -0.59
    Palest
    -0.59
     raped
    -0.57
    thood
    -0.57
    zbollah
    -0.57
    POSITIVE LOGITS
     this
    0.88
     these
    0.87
     Carney
    0.69
     the
    0.66
     both
    0.64
    these
    0.64
    :]
    0.63
     Larson
    0.61
     Gibbs
    0.59
     similarities
    0.59
    Activations Density 0.790%

    No Known Activations