© 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. 2946
    Prev
    Next
    INDEX
    Explanations

    mention of the name "Paul" with highest activation followed by the number 9

    oai_token-act-pair · gpt-3.5-turbo

    the name "Paul" in various contexts

    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
     lodge
    -0.74
    âĸ¬
    -0.71
     CONTROL
    -0.66
    UGE
    -0.64
    PRES
    -0.63
    GROUND
    -0.62
     Elven
    -0.62
    Limit
    -0.62
    ACTION
    -0.61
    âĸ¬âĸ¬
    -0.61
    POSITIVE LOGITS
    ains
    0.98
    son
    0.97
    ine
    0.96
     McCartney
    0.96
    sburg
    0.95
    ozo
    0.94
    okes
    0.93
    sen
    0.89
    sonian
    0.88
    anus
    0.87
    Activations Density 0.014%

    No Known Activations