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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 108
    Prev
    Next
    INDEX
    Explanations

    new information or reports

    oai_token-act-pair · gpt-3.5-turbo

    references to new reports or studies

    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
    cius
    -0.77
    cientious
    -0.76
    Zip
    -0.76
    ashtra
    -0.75
    uddin
    -0.74
    xual
    -0.72
    ufact
    -0.71
     actionGroup
    -0.70
    mop
    -0.68
    oleon
    -0.67
    POSITIVE LOGITS
    bie
    1.08
     arrivals
    1.02
    bies
    1.02
     additions
    1.01
     developments
    0.97
    foundland
    0.93
     revelations
    0.92
     YORK
    0.89
     discoveries
    0.85
     entrants
    0.83
    Activations Density 0.102%

    No Known Activations