© 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. 8-RES-JB
    11. 12404
    Prev
    Next
    INDEX
    Explanations

    phrases related to achieving a goal or making progress

    oai_token-act-pair · gpt-3.5-turbo

    the verb "make" and its various forms in different contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    bour
    -0.74
    Fram
    -0.73
    heter
    -0.69
    hood
    -0.60
    pac
    -0.60
     vest
    -0.60
     notwithstanding
    -0.59
    lex
    -0.59
    riger
    -0.59
     awaiting
    -0.59
    POSITIVE LOGITS
    hift
    1.04
     landfall
    0.95
     sure
    0.95
     headlines
    0.91
     excuses
    0.81
     strides
    0.79
     mistakes
    0.78
    netflix
    0.77
     sense
    0.76
    ailable
    0.76
    Activations Density 0.136%

    No Known Activations