© 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
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 3-RES-JB
    6. 312
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate a delay in time or action

    oai_token-act-pair · gpt-3.5-turbo

    references to the word "later."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    HO
    -0.72
    washing
    -0.71
    Ped
    -0.71
    HY
    -0.69
    FORE
    -0.67
    TRY
    -0.67
    condition
    -0.66
    owship
    -0.65
    EO
    -0.65
    tick
    -0.64
    POSITIVE LOGITS
     iterations
    0.95
     generations
    0.92
     versions
    0.89
     batches
    0.80
     phases
    0.78
     installments
    0.78
    noon
    0.78
     stages
    0.76
    aneously
    0.72
     than
    0.69
    Activations Density 0.032%

    No Known Activations