© 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. 8-RES-JB
    11. 16962
    Prev
    Next
    INDEX
    Explanations

    instances of legal or political terms

    oai_token-act-pair · gpt-3.5-turbo

    discussions surrounding race-related issues and their significance

    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
     Saiyan
    -0.64
     originals
    -0.63
     respectively
    -0.63
    album
    -0.63
     Rohing
    -0.62
    IZE
    -0.62
    UTF
    -0.62
     Sith
    -0.59
    \/\/
    -0.59
     successors
    -0.58
    POSITIVE LOGITS
     synonymous
    0.86
     legalized
    0.81
     compulsory
    0.76
    MODE
    0.72
     decriminal
    0.70
     sacrament
    0.69
     everywhere
    0.69
     predominant
    0.69
     trump
    0.69
    phy
    0.68
    Activations Density 0.669%

    No Known Activations