© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 2-RES-JB
    11. 7153
    Prev
    Next
    INDEX
    Explanations

    words related to the name "Aki"

    oai_token-act-pair · gpt-3.5-turbo

    mentions of specific individuals, particularly those with the suffix "aki"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    child
    -0.89
    actionDate
    -0.81
    advertisement
    -0.74
    cru
    -0.68
    craft
    -0.67
    sent
    -0.66
    spin
    -0.66
    aign
    -0.63
    neys
    -0.63
    APH
    -0.62
    POSITIVE LOGITS
    aki
    1.19
    uchi
    1.01
    wi
    0.92
    ya
    0.88
    veyard
    0.88
    unta
    0.87
    emon
    0.87
    adan
    0.85
    atsu
    0.85
    oba
    0.84
    Activations Density 0.007%

    No Known Activations