© 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. 8-RES-JB
    11. 8954
    Prev
    Next
    INDEX
    Explanations

    financial or numerical information expressed in a specific format

    oai_token-act-pair · gpt-3.5-turbo

    elements related to statistical and mathematical symbols or operations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    fires on tokens where the next token is not seperated by a space

    Explanation Uploaded by User
    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
    •Partition Features Layer 8 GPT2 Small
    " token" vs "token"

    No Comments

    Negative Logits
     metic
    -0.91
     Polk
    -0.85
     Union
    -0.85
     Dent
    -0.83
     Levy
    -0.82
     Grill
    -0.82
     Tier
    -0.82
     Bag
    -0.82
     Heroic
    -0.81
     XI
    -0.79
    POSITIVE LOGITS
    move
    1.59
    follow
    1.55
    subject
    1.55
    requ
    1.53
    when
    1.52
    cond
    1.52
    this
    1.51
    moving
    1.50
    sign
    1.50
    each
    1.50
    Activations Density 0.552%

    No Known Activations