© 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. 7-RES-JB
    11. 159
    Prev
    Next
    INDEX
    Explanations

    names of people who are speaking in a text

    oai_token-act-pair · gpt-3.5-turbo

    names followed by the verb "said" or "says" indicating attribution of statements or quotes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    7
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     retaliate
    -0.68
    venge
    -0.65
     clubhouse
    -0.62
     pageant
    -0.62
    itent
    -0.61
     revenge
    -0.60
     intimidate
    -0.60
    wcs
    -0.59
     audition
    -0.57
     lockout
    -0.57
    POSITIVE LOGITS
     recommends
    0.79
     concludes
    0.78
     et
    0.75
     argues
    0.75
     observes
    0.72
     warns
    0.71
     explains
    0.69
    gaard
    0.65
     gmaxwell
    0.64
     researcher
    0.63
    Activations Density 0.625%

    No Known Activations