INDEX

Explanations

summaries or recaps of information

oai_token-act-pair · gpt-3.5-turbo

phrases related to summaries and overviews

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

liam

-0.80

lees

-0.72

 Haram

-0.68

EED

-0.63

 Bronze

-0.60

yards

-0.59

existent

-0.59

oct

-0.59

obyl

-0.59

arsity

-0.59

POSITIVE LOGITS

 overview

0.95

 excerpt

0.93

 summar

0.84

 illustrating

0.81

helm

0.81

 refres

0.78

 explan

0.78

 highlighting

0.76

 summary

0.76

 synopsis

0.76

Activations Density 0.216%

summaries or recaps of information

phrases related to summaries and overviews

No Comments

No Known Activations