© 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. 5482
    Prev
    Next
    INDEX
    Explanations

    adjectives describing something as not needed or excessive

    oai_token-act-pair · gpt-3.5-turbo

    the term "unnecessary" used to describe actions, treatments, or situations

    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
    ebus
    -0.81
    ramid
    -0.80
    ois
    -0.79
    mitt
    -0.79
    eming
    -0.77
    king
    -0.74
    dar
    -0.74
    ker
    -0.74
    lining
    -0.72
    oin
    -0.72
    POSITIVE LOGITS
     unnecessary
    1.33
     needless
    1.03
     superflu
    1.03
     unnecess
    1.02
     duplication
    0.96
     complication
    0.89
     precaution
    0.86
     guiActiveUn
    0.84
     waste
    0.84
     pointless
    0.84
    Activations Density 0.008%

    No Known Activations