© 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 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. 3-RES-JB
    11. 4847
    Prev
    Next
    INDEX
    Explanations

    Roman numerals followed by a period

    oai_token-act-pair · gpt-3.5-turbo

    instances of the letter 'i'

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Valkyrie
    -0.76
     Wildcats
    -0.74
     Debor
    -0.69
    pter
    -0.67
    EStream
    -0.67
     Darwin
    -0.67
     Shelby
    -0.66
     colleg
    -0.64
     malaria
    -0.63
     Territ
    -0.62
    POSITIVE LOGITS
    ibo
    1.07
    ota
    1.05
    ñ
    1.02
    wa
    0.98
    pec
    0.97
    owa
    0.96
    orno
    0.96
    oga
    0.93
    uli
    0.93
    abor
    0.92
    Activations Density 0.040%

    No Known Activations