INDEX

Explanations

phrases indicating differentiating or categorizing words or concepts

oai_token-act-pair · gpt-3.5-turbo

repeated phrases or quotations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ticket

-0.76

 periodic

-0.75

 bending

-0.74

 targeted

-0.73

 undocumented

-0.72

 dipping

-0.72

 whistle

-0.72

 rall

-0.71

 pacing

-0.71

oun

-0.71

POSITIVE LOGITS

soDeliveryDate

1.42

short

1.21

true

1.20

Si

1.13

Nature

1.12

false

1.11

nothing

1.10

inventoryQuantity

1.10

little

1.08

final

1.07

Activations Density 0.070%

phrases indicating differentiating or categorizing words or concepts

repeated phrases or quotations

No Comments

No Known Activations