© 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. 0-RES-JB
    11. 1092
    Prev
    Next
    INDEX
    Explanations

    mentions of drug overdoses

    oai_token-act-pair · gpt-3.5-turbo

    terms related to drug overdoses and poisoning

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Kem
    -0.73
    dar
    -0.72
     sab
    -0.71
    å§
    -0.70
     polit
    -0.69
     Mach
    -0.68
    ERA
    -0.68
    roud
    -0.68
     MIL
    -0.68
     Franchise
    -0.68
    POSITIVE LOGITS
     overdose
    3.44
     overdoses
    2.97
     poisoning
    2.00
    dose
    1.86
    Benz
    1.81
    entanyl
    1.77
     intoxication
    1.66
     fentanyl
    1.60
     overd
    1.59
     toxicity
    1.53
    Activations Density 0.038%

    No Known Activations