INDEX

Explanations

words related to ears or ear-related objects

oai_token-act-pair · gpt-3.5-turbo

references to ears and auditory elements 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 @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ERO

-0.78

VK

-0.76

 Alam

-0.73

ABE

-0.72

 Vide

-0.71

antine

-0.65

 Commonwealth

-0.64

enance

-0.64

 Williamson

-0.63

 Exec

-0.63

POSITIVE LOGITS

phones

1.25

ldom

1.25

rings

1.17

 lobe

1.16

phone

1.12

piece

1.11

 canal

1.06

wig

1.04

marked

1.03

pieces

0.99

Activations Density 0.030%

words related to ears or ear-related objects

references to ears and auditory elements in various contexts

No Comments

No Known Activations