© 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. 6594
    Prev
    Next
    INDEX
    Explanations

    negative concepts or entities often associated with antagonists or conflicts, such as 'evil'

    oai_token-act-pair · gpt-3.5-turbo

    references to the concept of evil

    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
    UNCH
    -0.76
    raltar
    -0.75
    PsyNetMessage
    -0.74
    ribe
    -0.72
    Lago
    -0.71
    dropping
    -0.71
    illation
    -0.71
    hene
    -0.70
    aro
    -0.70
    akeru
    -0.70
    POSITIVE LOGITS
     incarn
    0.98
     deeds
    0.92
     deed
    0.86
     genius
    0.84
     undermin
    0.82
    nesses
    0.82
     lord
    0.81
     evil
    0.81
     mastermind
    0.80
     villain
    0.78
    Activations Density 0.015%

    No Known Activations