INDEX

Explanations

words related to physical actions, intense emotions, and descriptions of scenes

oai_token-act-pair · gpt-3.5-turbo

punctuation and its patterns within the text

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

OND

-0.77

UF

-0.74

venth

-0.68

":"/

-0.66

emaker

-0.66

ONSORED

-0.66

":[{"

-0.65

®

-0.65

enary

-0.65

pec

-0.64

POSITIVE LOGITS

 albeit

1.03

 culminating

0.92

 lest

0.91

 including

0.83

huh

0.79

 alas

0.79

 namely

0.77

 though

0.75

 plung

0.74

eh

0.74

Activations Density 0.832%

words related to physical actions, intense emotions, and descriptions of scenes

punctuation and its patterns within the text

No Comments

No Known Activations