© 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. 218
    Prev
    Next
    INDEX
    Explanations

    the word "capital" when referring to cities or locations

    oai_token-act-pair · gpt-3.5-turbo

    references to various political or geographical capitals

    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
    AUT
    -0.88
    sbm
    -0.86
    ĪĴ
    -0.81
    »Ĵ
    -0.75
    lyak
    -0.73
    chrom
    -0.70
    potion
    -0.70
    fortune
    -0.69
    MpServer
    -0.69
    hner
    -0.68
    POSITIVE LOGITS
     metropolitan
    0.88
     suburb
    0.84
     punishment
    0.80
     capital
    0.76
     city
    0.76
    orpor
    0.76
    itals
    0.75
    omach
    0.74
    uania
    0.71
     cities
    0.69
    Activations Density 0.013%

    No Known Activations