© 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 EvalsExportsinterp-engineNEWAPI 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. 9-RES-JB
    11. 23429
    Prev
    Next
    INDEX
    Explanations

    phrases related to family members and their actions or attributes

    oai_token-act-pair · gpt-3.5-turbo

    commas and other punctuation marks that typically accompany names or phrases

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    9
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ãĥ¥
    -0.66
    ishers
    -0.63
     Compatibility
    -0.60
    mopolitan
    -0.59
    rolley
    -0.57
    ESA
    -0.55
    clusive
    -0.55
    Different
    -0.54
     Hearth
    -0.54
    ãĤ¤
    -0.54
    POSITIVE LOGITS
     Jr
    1.01
     aka
    0.90
     who
    0.82
     PhD
    0.79
     withdrew
    0.78
     disagrees
    0.78
    who
    0.77
     whose
    0.75
     denies
    0.75
     insists
    0.74
    Activations Density 0.145%

    No Known Activations