INDEX

Explanations

phrases related to quotes or interviews, specifically sentences beginning with quotation marks or involving statements by individuals

oai_token-act-pair · gpt-3.5-turbo

expressions of emotional intensity or emphasis

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

 shroud

-0.85

 shrouded

-0.76

 semblance

-0.76

 presumed

-0.68

 Mayo

-0.66

 grips

-0.64

 Mirage

-0.63

interstitial

-0.61

 slate

-0.61

 wrapped

-0.61

POSITIVE LOGITS

¹

1.23

ł

1.16

Ķ

0.98

ª

0.95

¾

0.94

ĳ

0.93

ı

0.92

wcsstore

0.92

»

0.92

³

0.92

Activations Density 0.180%

phrases related to quotes or interviews, specifically sentences beginning with quotation marks or involving statements by individuals

expressions of emotional intensity or emphasis

No Comments

No Known Activations