INDEX

Explanations

phrases related to expressing personal viewpoints or beliefs

oai_token-act-pair · gpt-3.5-turbo

mentions of personal opinions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Gutenberg

-0.78

gm

-0.68

 Roose

-0.67

ammers

-0.66

artifacts

-0.66

ammy

-0.65

 Mamm

-0.64

artney

-0.64

estones

-0.63

 Danger

-0.63

POSITIVE LOGITS

 opinion

0.90

 opinions

0.87

 largeDownload

0.82

 piece

0.76

 polls

0.75

 opin

0.74

atively

0.73

 bias

0.72

eering

0.72

 paralysis

0.70

Activations Density 0.020%

phrases related to expressing personal viewpoints or beliefs

mentions of personal opinions

No Comments

No Known Activations