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

    Neuronpedia

    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. 8-RES-JB
    11. 8407
    Prev
    Next
    INDEX
    Explanations

    references to physical or architectural structures such as monuments or buildings

    oai_token-act-pair · gpt-3.5-turbo

    references to money or monetary concepts

    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
     Norn
    -0.86
    ORTS
    -0.75
    FUL
    -0.75
    INGTON
    -0.74
    IBLE
    -0.69
    UGE
    -0.69
    ULTS
    -0.68
     Venom
    -0.68
     lockout
    -0.67
     enrichment
    -0.65
    POSITIVE LOGITS
    stros
    1.47
    olithic
    1.34
    olith
    1.29
    soon
    1.26
    ograph
    1.20
    ocular
    1.17
    ographs
    1.12
    ologue
    1.11
    oton
    1.08
    omial
    1.07
    Activations Density 0.017%

    No Known Activations