© 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. 8-RES-JB
    11. 9811
    Prev
    Next
    INDEX
    Explanations

    instances of people meeting each other

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "met" and its variations, indicating encounters or meetings

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     untreated
    -0.67
    vernment
    -0.67
     membranes
    -0.66
     upkeep
    -0.65
    secut
    -0.64
     incarcer
    -0.64
    otine
    -0.64
     conce
    -0.63
    inth
    -0.63
    mort
    -0.62
    POSITIVE LOGITS
    amorph
    1.17
    ropolis
    0.89
    agame
    0.78
    ioned
    0.77
    ered
    0.75
     fellow
    0.74
    atron
    0.72
    lehem
    0.71
    tle
    0.71
     collaborators
    0.71
    Activations Density 0.043%

    No Known Activations