© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 11-RES-JB
    11. 6327
    Prev
    Next
    INDEX
    Explanations

    numbers and statistics related to various trends and changes in data

    oai_token-act-pair · gpt-3.5-turbo

    quantitative descriptions and statistics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Edit
    -0.78
    inside
    -0.76
    words
    -0.72
    speak
    -0.72
    ultimate
    -0.72
    ographs
    -0.71
    iar
    -0.70
    Area
    -0.70
    istar
    -0.69
     aliases
    -0.68
    POSITIVE LOGITS
     decrease
    1.48
     reduction
    1.32
     slowdown
    1.27
     spike
    1.23
     decline
    1.23
     surge
    1.22
     rise
    1.18
     gradual
    1.18
     deterioration
    1.17
     doubling
    1.17
    Activations Density 0.200%

    No Known Activations