© 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. 1802
    Prev
    Next
    INDEX
    Explanations

    phrases related to instances of drunk driving

    oai_token-act-pair · gpt-3.5-turbo

    references to alcohol consumption and driving under the influence

    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
     Flavoring
    -0.91
    pha
    -0.72
     JPM
    -0.71
     Hosp
    -0.71
    ocol
    -0.68
     Folder
    -0.66
    Downloadha
    -0.66
    ":["
    -0.65
    metics
    -0.64
    CVE
    -0.64
    POSITIVE LOGITS
    ards
    0.95
     manslaughter
    0.91
    driving
    0.89
     drunk
    0.89
    ard
    0.84
     drinking
    0.83
    bott
    0.82
     driving
    0.80
     underage
    0.80
    cohol
    0.78
    Activations Density 0.051%

    No Known Activations