© 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. 8-RES-JB
    11. 2673
    Prev
    Next
    INDEX
    Explanations

    phrases related to accomplishments or achievements

    oai_token-act-pair · gpt-3.5-turbo

    the word "has" and its variations in different contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     drafting
    -0.59
     behold
    -0.58
     territ
    -0.58
     MacArthur
    -0.57
    oppable
    -0.55
     unaccount
    -0.55
    ecd
    -0.54
     paycheck
    -0.54
     Allied
    -0.54
     persuasion
    -0.53
    POSITIVE LOGITS
     been
    1.13
    been
    1.05
    bara
    0.99
     Been
    0.90
     undergone
    0.90
    hed
    0.88
     emerged
    0.88
    kell
    0.86
     begun
    0.84
     become
    0.84
    Activations Density 0.077%

    No Known Activations