© 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. 9608
    Prev
    Next
    INDEX
    Explanations

    references to a specific timeframe, specifically the word "decade" either in singular or plural form

    oai_token-act-pair · gpt-3.5-turbo

    references to the concept of "decades" or specific time periods

    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
    anamo
    -0.85
    iquid
    -0.74
    ournals
    -0.74
    iqu
    -0.71
    govtrack
    -0.69
    ettings
    -0.68
     ingred
    -0.67
    ffic
    -0.67
    pots
    -0.67
    ktop
    -0.66
    POSITIVE LOGITS
     ago
    0.92
     ILCS
    0.81
    ĸļ
    0.77
     Ago
    0.75
    icity
    0.71
     olds
    0.71
     anniversary
    0.69
     Anniversary
    0.69
     onward
    0.69
    uously
    0.68
    Activations Density 0.028%

    No Known Activations