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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 1-RES-JB
    11. 16335
    Prev
    Next
    INDEX
    Explanations

    names of newspapers

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "Globe" and references to the region of Newfoundland

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ramid
    -0.80
    hyde
    -0.78
    araoh
    -0.74
    ded
    -0.74
    tes
    -0.73
    sen
    -0.73
    amac
    -0.72
    idy
    -0.72
    tan
    -0.72
    paces
    -0.72
    POSITIVE LOGITS
     Globe
    1.38
     Herald
    0.94
     Newspaper
    0.89
     Scotia
    0.82
     Spotlight
    0.81
     Harbour
    0.78
     Citizen
    0.72
     Defender
    0.70
     Journalism
    0.69
     Awards
    0.69
    Activations Density 0.003%

    No Known Activations