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

    Neuronpedia

    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. 4-RES-JB
    11. 14505
    Prev
    Next
    INDEX
    Explanations

    mentions or variations of the word "Bend"

    oai_token-act-pair · gpt-3.5-turbo

    references to specific locations, particularly named places

    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
     readable
    -0.70
     dramas
    -0.69
     spoken
    -0.68
     conclud
    -0.66
    arily
    -0.65
     handsome
    -0.64
     sophisticated
    -0.64
     Carbuncle
    -0.63
     drama
    -0.63
     pricey
    -0.62
    POSITIVE LOGITS
     Bend
    1.12
    SourceFile
    0.94
    owski
    0.86
    ford
    0.84
    stead
    0.82
    mire
    0.82
    eret
    0.82
    ola
    0.82
    pole
    0.81
    luck
    0.80
    Activations Density 0.011%

    No Known Activations