INDEX

Explanations

places or entities associated with a location

oai_token-act-pair · gpt-3.5-turbo

words and phrases related to government entities and their potential actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Wall

-0.77

 Horn

-0.74

Hill

-0.74

 Wall

-0.72

 Kear

-0.70

 Page

-0.69

ocratic

-0.67

 Trail

-0.66

 Hammer

-0.64

Klu

-0.64

POSITIVE LOGITS

fect

4.33

fecture

3.25

duction

1.07

intend

0.98

merga

0.96

aware

0.96

genic

0.96

pora

0.95

infect

0.91

uctor

0.89

Activations Density 0.039%

places or entities associated with a location

words and phrases related to government entities and their potential actions

No Comments

No Known Activations