INDEX

Explanations

phrases related to authenticity or credibility

oai_token-act-pair · gpt-3.5-turbo

terms indicating authenticity or truthfulness

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

trap

-0.74

 Slate

-0.73

 Twins

-0.71

 Cull

-0.69

Keys

-0.68

 Sparrow

-0.68

 Prism

-0.66

 Marshal

-0.66

 Daughter

-0.65

Orn

-0.62

POSITIVE LOGITS

 fide

1.08

 bona

0.98

 factual

0.91

ual

0.89

uine

0.85

ulty

0.85

oreal

0.80

iological

0.77

 basis

0.77

ysical

0.76

Activations Density 0.025%

phrases related to authenticity or credibility

terms indicating authenticity or truthfulness

No Comments

No Known Activations