© 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. 2-RES-JB
    11. 13603
    Prev
    Next
    INDEX
    Explanations

    arguments or points being made in a text

    oai_token-act-pair · gpt-3.5-turbo

    references to arguments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     stocking
    -0.84
    annis
    -0.72
     livest
    -0.70
    eco
    -0.69
    oho
    -0.69
    pleting
    -0.68
    cy
    -0.67
    cler
    -0.67
    hew
    -0.67
     Seym
    -0.66
    POSITIVE LOGITS
     arguments
    1.11
     argument
    1.05
    uments
    1.04
    argument
    0.95
    edIn
    0.90
    -+-+
    0.90
    ative
    0.85
    ļéĨĴ
    0.81
    @@@@@@@@
    0.79
    Cooldown
    0.79
    Activations Density 0.019%

    No Known Activations