© 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. 7793
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal charges and criminal activities

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to criminal charges

    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
    BUG
    -0.66
    clips
    -0.65
    phrine
    -0.64
    bytes
    -0.64
     periphery
    -0.64
     contrace
    -0.64
    Sport
    -0.62
    Tokens
    -0.61
    ats
    -0.61
    chan
    -0.61
    POSITIVE LOGITS
     regards
    0.89
    holding
    0.83
     regard
    0.82
    regon
    0.82
     respect
    0.80
    standing
    0.78
    stood
    0.77
     felony
    0.74
     upholding
    0.74
     treason
    0.73
    Activations Density 0.048%

    No Known Activations