INDEX

Explanations

words related to governments or political entities

oai_token-act-pair · gpt-3.5-turbo

words that contain the letter 'e' in varying contexts

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

hyde

-1.02

Ĥİ

-0.77

terday

-0.76

etheless

-0.76

 Annotations

-0.75

å§«

-0.75

ĸļ

-0.74

 Macedonia

-0.72

 Fargo

-0.72

 Rica

-0.72

POSITIVE LOGITS

iest

1.19

erb

1.14

ashing

1.09

est

1.09

ounding

1.00

uper

1.00

ashed

0.99

amen

0.99

attered

0.98

agging

0.97

Activations Density 0.350%

words related to governments or political entities

words that contain the letter 'e' in varying contexts

No Comments

No Known Activations