© 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. 8-RES-JB
    11. 3166
    Prev
    Next
    INDEX
    Explanations

    specific time references, particularly in the evening, frequently mentioned in TV schedule descriptions

    oai_token-act-pair · gpt-3.5-turbo

    references to specific times of day

    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
     acknowled
    -0.68
     resemb
    -0.68
    govern
    -0.65
     heirs
    -0.63
    ollah
    -0.62
     behav
    -0.62
    Bey
    -0.60
    llah
    -0.59
    smanship
    -0.59
    beit
    -0.58
    POSITIVE LOGITS
     EDT
    1.27
     PDT
    1.25
     EST
    1.23
     PST
    1.21
     CET
    1.12
     GMT
    1.11
     CST
    1.09
     BST
    1.06
     ET
    1.02
     UTC
    0.96
    Activations Density 0.034%

    No Known Activations