INDEX

Explanations

numerical values embedded in text strings

oai_token-act-pair · gpt-3.5-turbo

numerical data or statistics related to discussions or reports

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

awaru

-0.70

rave

-0.66

hold

-0.63

Thumbnail

-0.62

 latter

-0.62

arial

-0.61

ably

-0.61

ial

-0.59

 bother

-0.58

 proportion

-0.58

POSITIVE LOGITS

hrs

0.98

 Months

0.95

 Hours

0.93

 mins

0.92

 IMAGES

0.91

ONSORED

0.91

 Minutes

0.84

 Years

0.82

ãĤ¼ãĤ¦ãĤ¹

0.82

 âĺħ

0.82

Activations Density 0.105%

numerical values embedded in text strings

numerical data or statistics related to discussions or reports

No Comments

No Known Activations