INDEX

Explanations

names of specific individuals, particularly "Logan" and "Wolverine."

oai_token-act-pair · gpt-3.5-turbo

mentions of the names "Logan" and "Wolverine."

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

iculty

-0.72

ible

-0.68

chool

-0.67

establishment

-0.67

earance

-0.66

essler

-0.65

guided

-0.65

apers

-0.64

ysical

-0.64

FER

-0.64

POSITIVE LOGITS

culosis

0.92

istani

0.87

 Logan

0.79

ized

0.76

ty

0.75

isation

0.75

mire

0.74

lain

0.74

ously

0.73

ization

0.73

Activations Density 0.167%

names of specific individuals, particularly "Logan" and "Wolverine."

mentions of the names "Logan" and "Wolverine."

No Comments

No Known Activations