INDEX

Explanations

instances of the word "every" followed by various context clues

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "every" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

redients

-0.83

ammers

-0.80

olas

-0.79

Rs

-0.77

ees

-0.76

loads

-0.72

 Presents

-0.68

igers

-0.68

ptoms

-0.68

holders

-0.67

POSITIVE LOGITS

 conceivable

1.68

 imaginable

1.57

 facet

1.24

 aspect

1.22

THING

1.20

 single

1.05

 possible

1.00

 major

0.95

 corner

0.95

 inch

0.94

Activations Density 0.060%

instances of the word "every" followed by various context clues

occurrences of the word "every" in various contexts

No Comments

No Known Activations