INDEX

Explanations

phrases related to academic or scholarly contexts

oai_token-act-pair · gpt-3.5-turbo

references to scientific institutions, publications, and fields

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

plet

-0.80

 glim

-0.72

yond

-0.69

rises

-0.68

ãĥİ

-0.68

 promoters

-0.65

rise

-0.65

BUS

-0.64

cells

-0.63

details

-0.63

POSITIVE LOGITS

Enlarge

0.82

 Costume

0.80

ingham

0.75

 Foundation

0.74

 Appeals

0.73

 Conference

0.73

Newsletter

0.72

 Mellon

0.70

 Examination

0.67

ozo

0.67

Activations Density 0.086%

phrases related to academic or scholarly contexts

references to scientific institutions, publications, and fields

No Comments

No Known Activations