INDEX

Explanations

phrases related to information being heard or heard about

oai_token-act-pair · gpt-3.5-turbo

phrases related to the concept of hearing or being informed

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

PHOTOS

-0.74

ahime

-0.70

sis

-0.70

ila

-0.69

igsaw

-0.69

ãĤ¦

-0.68

tan

-0.67

iasis

-0.65

eros

-0.64

els

-0.64

POSITIVE LOGITS

 voices

1.29

 footsteps

1.01

 confessions

0.99

 whispers

0.95

 rumours

0.92

 noises

0.89

 Voices

0.88

 rumors

0.88

 testimonies

0.86

 loud

0.85

Activations Density 0.058%

phrases related to information being heard or heard about

phrases related to the concept of hearing or being informed

No Comments

No Known Activations