© 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. 3-RES-JB
    11. 4803
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "Tah" followed by a combination of letters

    oai_token-act-pair · gpt-3.5-turbo

    references to Tahir Square

    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
    mble
    -0.75
    ablishment
    -0.74
    ISTER
    -0.73
     fixation
    -0.71
     cov
    -0.69
    ancial
    -0.61
     Parenthood
    -0.61
    ãģķ
    -0.60
     bleed
    -0.60
     psy
    -0.60
    POSITIVE LOGITS
     Tah
    1.06
    oya
    1.02
    rir
    1.02
    ibia
    0.88
    essa
    0.88
    una
    0.87
    iri
    0.86
    anus
    0.86
    ti
    0.86
    iti
    0.85
    Activations Density 0.007%

    No Known Activations