© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 24372
    Prev
    Next
    INDEX
    Explanations

    legal and news-related terms and proper nouns such as "Copyright", "News", and "shooter"

    oai_token-act-pair · gpt-3.5-turbo

    mentions of copyright and associated legal terms

    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
    itance
    -0.76
    arial
    -0.74
    utor
    -0.74
    table
    -0.73
    nikov
    -0.70
    pora
    -0.69
    ometric
    -0.68
     stockpile
    -0.68
    ndra
    -0.68
    ges
    -0.67
    POSITIVE LOGITS
    radio
    1.11
     radio
    1.08
    Radio
    1.03
     Radio
    1.01
    CBS
    0.94
    TV
    0.90
    Channel
    0.89
     News
    0.87
     broadcasts
    0.87
     Asheville
    0.86
    Activations Density 0.128%

    No Known Activations