INDEX

Explanations

phrases related to lists of items or categories

oai_token-act-pair · gpt-3.5-turbo

phrases indicating quantities or amounts, often emphasizing the word "more."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

lance

-0.87

Joy

-0.73

Unity

-0.70

ivism

-0.69

stadt

-0.68

heed

-0.68

POST

-0.67

BIL

-0.66

rix

-0.66

gow

-0.66

POSITIVE LOGITS

 dozen

1.13

 hundred

0.95

 paragraphs

0.93

 consecutive

0.93

 sectors

0.91

 layers

0.91

 thousand

0.90

 episodes

0.89

 segments

0.89

 exceptions

0.88

Activations Density 0.047%

phrases related to lists of items or categories

phrases indicating quantities or amounts, often emphasizing the word "more."

No Comments

No Known Activations