© 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. 0-RES-JB
    11. 2153
    Prev
    Next
    INDEX
    Explanations

    phrases including the word "row"

    oai_token-act-pair · gpt-3.5-turbo

    references to rows in a structured format, likely indicating table-like data or bullet points

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     authentic
    -0.74
     authenticity
    -0.70
     imitation
    -0.70
     Esper
    -0.69
     perfection
    -0.65
     Amit
    -0.64
     Tas
    -0.61
     aspir
    -0.61
     BIOS
    -0.61
     antit
    -0.60
    POSITIVE LOGITS
    row
    4.69
    rows
    3.02
    rower
    2.42
    rowing
    1.84
    rown
    1.56
    Row
    1.55
    rew
    1.37
     row
    1.36
     Row
    1.23
     rows
    1.20
    Activations Density 0.013%

    No Known Activations