INDEX

Explanations

adjectives describing subpar or negative quality

oai_token-act-pair · gpt-3.5-turbo

terms related to poor quality or stagnant performance

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

dar

-0.77

raltar

-0.74

frey

-0.73

saw

-0.72

thouse

-0.71

Ru

-0.69

leans

-0.68

ordering

-0.67

larg

-0.66

arta

-0.66

POSITIVE LOGITS

 miser

0.88

luster

0.87

 mediocre

0.68

nesses

0.68

 misc

0.66

 Yamato

0.66

 grades

0.66

 Downs

0.65

igue

0.65

 excuses

0.65

Activations Density 0.065%

adjectives describing subpar or negative quality

terms related to poor quality or stagnant performance

No Comments

No Known Activations