© 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. 3-RES-JB
    11. 14795
    Prev
    Next
    INDEX
    Explanations

    verbs or phrases related to things being split or divided

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "separated" and its variations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    vous
    -0.71
    ×Ķ
    -0.70
    WN
    -0.67
    OD
    -0.67
    cycl
    -0.65
    TL
    -0.65
    nz
    -0.65
     Briggs
    -0.63
     Gos
    -0.63
    enos
    -0.63
    POSITIVE LOGITS
     separating
    0.91
    separ
    0.84
     separ
    0.83
     separated
    0.83
     sexes
    0.81
     separates
    0.76
     detach
    0.74
    ĨĴ
    0.73
    ively
    0.73
    icut
    0.73
    Activations Density 0.017%

    No Known Activations