INDEX

Explanations

mentions of specific names, phrases, or terms

oai_token-act-pair · gpt-3.5-turbo

instances of the word "mentioned."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

earned

-0.74

isine

-0.72

usterity

-0.70

uilt

-0.69

orously

-0.69

 hearts

-0.68

heres

-0.68

edom

-0.67

 heart

-0.66

otypes

-0.66

POSITIVE LOGITS

 mentioning

1.12

 mentioned

1.07

 mentions

1.00

 mention

0.93

ãĤ®

0.78

 aloud

0.78

 above

0.76

REDACTED

0.76

 herein

0.75

lihood

0.71

Activations Density 0.008%

mentions of specific names, phrases, or terms

instances of the word "mentioned."

No Comments

No Known Activations