© 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. 20797
    Prev
    Next
    INDEX
    Explanations

    phrases related to illicit or dishonest activities

    oai_token-act-pair · gpt-3.5-turbo

    variations of the word "got" and phrases indicating possession or acquisition

    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
     challeng
    -0.79
     corrid
    -0.74
     livest
    -0.70
    unctions
    -0.68
     patents
    -0.68
    ĺħ
    -0.67
     cured
    -0.65
     tremend
    -0.65
     autonom
    -0.64
    uku
    -0.64
    POSITIVE LOGITS
    gotten
    1.36
    getting
    0.91
    give
    0.84
     Flavoring
    0.77
    fare
    0.77
    get
    0.73
     Macedonia
    0.73
    berus
    0.72
    nel
    0.72
    pret
    0.70
    Activations Density 0.018%

    No Known Activations