INDEX

Explanations

mentions of specific formatting elements within a text, such as advertisement separators or story continuations

oai_token-act-pair · gpt-3.5-turbo

instances of advertisement or promotional content

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

 homebrew

-0.73

addin

-0.65

EStream

-0.64

romy

-0.62

leans

-0.61

 adjunct

-0.60

boro

-0.60

ktop

-0.59

 estranged

-0.56

issance

-0.54

POSITIVE LOGITS

ccording

0.84

JUST

0.78

SPONSORED

0.72

Spain

0.71

ATT

0.67

AIN

0.67

eria

0.67

TAG

0.66

Reward

0.65

Prev

0.65

Activations Density 0.073%

mentions of specific formatting elements within a text, such as advertisement separators or story continuations

instances of advertisement or promotional content

No Comments

No Known Activations