© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 2-RES-JB
    11. 21255
    Prev
    Next
    INDEX
    Explanations

    negative actions or behaviors aimed at discrediting or harming someone

    oai_token-act-pair · gpt-3.5-turbo

    terms related to defamation and negative characterizations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    changes
    -0.82
    edience
    -0.81
    hang
    -0.78
    alted
    -0.77
    odes
    -0.76
    ECK
    -0.70
    akening
    -0.70
    estamp
    -0.69
    edient
    -0.69
    peror
    -0.69
    POSITIVE LOGITS
     smear
    1.00
     vil
    0.86
     slurs
    0.85
     slander
    0.80
     allegations
    0.80
     accusations
    0.74
    ãĥ¼ãĥĨãĤ£
    0.72
    ãĤ¹ãĥĪ
    0.71
     stereotypes
    0.69
    ãĥķãĤ¡
    0.68
    Activations Density 0.020%

    No Known Activations