© 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. 16407
    Prev
    Next
    INDEX
    Explanations

    phrases related to different types of relationships

    oai_token-act-pair · gpt-3.5-turbo

    phrases that mention various forms of relationships

    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
    dm
    -0.77
    hett
    -0.72
    stad
    -0.72
    ateurs
    -0.69
    ESH
    -0.69
    stall
    -0.67
    oggles
    -0.66
    gob
    -0.66
    asks
    -0.65
    boards
    -0.64
    POSITIVE LOGITS
     relationship
    3.65
     Relationship
    2.90
     relationships
    2.72
    relations
    2.32
     relation
    2.21
    Relations
    1.95
     relations
    1.94
     Relations
    1.86
     friendship
    1.79
     partnership
    1.73
    Activations Density 0.014%

    No Known Activations