INDEX

Explanations

references to a specific entity or term "WP" with varying activations

oai_token-act-pair · gpt-3.5-turbo

references to a specific entity or group identified as "WP"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

thus

-0.83

é¾įåĸļå£«

-0.82

ante

-0.82

Reviewer

-0.79

 thous

-0.75

hips

-0.74

erald

-0.73

taboola

-0.72

angelo

-0.72

tes

-0.71

POSITIVE LOGITS

WP

1.29

WP

1.28

olicy

1.05

Beg

0.81

FFER

0.74

witz

0.73

wordpress

0.72

LP

0.71

ITCH

0.71

ctive

0.71

Activations Density 0.005%

references to a specific entity or term "WP" with varying activations

references to a specific entity or group identified as "WP"

No Comments

No Known Activations