© 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. 2-RES-JB
    11. 23048
    Prev
    Next
    INDEX
    Explanations

    keywords related to substances, especially in the context of abuse or regulations

    oai_token-act-pair · gpt-3.5-turbo

    references to the word "substance" and related terms indicating types of substances

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    DIR
    -0.71
    opp
    -0.67
     Prophet
    -0.66
    aird
    -0.65
    athered
    -0.65
    addafi
    -0.64
    rium
    -0.63
    Rated
    -0.62
     footh
    -0.62
     Oracle
    -0.62
    POSITIVE LOGITS
     substance
    1.16
     Substance
    1.07
    fulness
    1.01
     substances
    0.89
    cohol
    0.81
    ulkan
    0.81
    amphetamine
    0.80
    estinal
    0.80
     abuser
    0.80
     ecstasy
    0.79
    Activations Density 0.006%

    No Known Activations