© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 12817
    Prev
    Next
    INDEX
    Explanations

    phrases related to physical location or spatial description

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating quantity and composition

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Negative Logits
    bowl
    -0.83
    tesy
    -0.81
    ivals
    -0.81
    initions
    -0.79
    acus
    -0.74
    isons
    -0.74
    idays
    -0.73
    days
    -0.72
    inations
    -0.72
    dose
    -0.72
    POSITIVE LOGITS
     flesh
    1.08
     ice
    1.02
     flats
    1.00
     grass
    0.98
     dough
    0.96
     parchment
    0.94
     rubble
    0.92
     asphalt
    0.91
     metal
    0.91
     molten
    0.91
    Activations Density 0.113%

    No Known Activations