© 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. 488
    Prev
    Next
    INDEX
    Explanations

    references to specific types of groups or organizations, particularly ones related to different sectors or belief systems

    oai_token-act-pair · gpt-3.5-turbo

    references to religious or spiritual groups

    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
     balance
    -0.73
    aterasu
    -0.70
    apeake
    -0.68
     meal
    -0.65
    balance
    -0.61
    ulet
    -0.61
    course
    -0.60
    aroo
    -0.59
     caffe
    -0.59
    ishable
    -0.59
    POSITIVE LOGITS
     Sect
    1.90
     sect
    1.64
     sects
    1.04
    )."
    0.74
    mble
    0.71
    Prev
    0.70
    '."
    0.70
     Sector
    0.69
    cill
    0.68
    recy
    0.64
    Activations Density 0.001%

    No Known Activations