© 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. 3-RES-JB
    11. 11414
    Prev
    Next
    INDEX
    Explanations

    terms related to geographical locations or entities with 'Ur' in them

    oai_token-act-pair · gpt-3.5-turbo

    repeated mentions of a specific location or entity, likely "Urban"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    女
    -0.80
    washer
    -0.80
    obyl
    -0.75
    iculty
    -0.72
    ãģ¦
    -0.72
    ãĥ¯
    -0.69
    pmwiki
    -0.68
     Sins
    -0.67
     sund
    -0.67
    æ³
    -0.67
    POSITIVE LOGITS
    gent
    1.10
    gently
    0.93
    gencies
    0.91
    gery
    0.89
    leys
    0.88
    du
    0.83
    seless
    0.82
    verse
    0.82
    anyl
    0.81
    pee
    0.80
    Activations Density 0.007%

    No Known Activations