© 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. 5630
    Prev
    Next
    INDEX
    Explanations

    references to problematic or negative situations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    strong profanity related to negative experiences or circumstances

    oai_token-act-pair · gpt-3.5-turbo
    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
    gary
    -0.77
    AUT
    -0.75
    ————————————————
    -0.68
    fman
    -0.66
    HCR
    -0.64
    âĢ¢âĢ¢âĢ¢âĢ¢
    -0.64
    rouse
    -0.64
    lique
    -0.64
    eur
    -0.63
    isconsin
    -0.63
    POSITIVE LOGITS
     shit
    1.08
    shit
    1.06
     Shit
    1.01
     crap
    0.96
    bags
    0.83
    dump
    0.80
    loads
    0.80
    fuck
    0.79
     fuck
    0.76
     heap
    0.74
    Activations Density 0.005%

    No Known Activations