© 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. 6417
    Prev
    Next
    INDEX
    Explanations

    mentions of time periods, especially "night", and quantitative descriptions, particularly related to cost and quality

    oai_token-act-pair · gpt-3.5-turbo

    terms associated with time and measurement, particularly in late-night contexts

    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
    angers
    -0.74
    iph
    -0.68
    seys
    -0.67
    phies
    -0.65
    inges
    -0.64
    ktop
    -0.64
    erves
    -0.63
    EStream
    -0.62
    aps
    -0.62
    agi
    -0.61
    POSITIVE LOGITS
    tenance
    1.16
    theless
    0.92
     survival
    0.80
    etheless
    0.75
    backer
    0.72
    lihood
    0.72
    osterone
    0.71
    ruction
    0.71
     correspondence
    0.71
    istance
    0.67
    Activations Density 0.168%

    No Known Activations