© 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. 8-RES-JB
    11. 11230
    Prev
    Next
    INDEX
    Explanations

    phrases related to providing additional details or background information

    oai_token-act-pair · gpt-3.5-turbo

    statements that convey concluding thoughts or summarizations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
    Not in Any Lists

    No Comments

    Negative Logits
     condol
    -0.71
     bul
    -0.68
     appro
    -0.67
     sidelines
    -0.67
     integ
    -0.66
     minim
    -0.66
     unbeat
    -0.66
     gravy
    -0.64
     impatient
    -0.64
     aval
    -0.64
    POSITIVE LOGITS
     Firstly
    1.65
     Suppose
    1.58
     Whereas
    1.29
     Whenever
    1.26
     Whilst
    1.24
     When
    1.20
     First
    1.15
     Normally
    1.13
     Because
    1.13
     While
    1.12
    Activations Density 0.360%

    No Known Activations