INDEX

Explanations

words related to taxation and divisive or separating elements

oai_token-act-pair · gpt-3.5-turbo

language related to divisive issues or taxes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ession

-0.83

kins

-0.83

ainer

-0.77

esses

-0.75

ystem

-0.75

hips

-0.74

grounds

-0.72

icht

-0.71

 Chaser

-0.70

kin

-0.70

POSITIVE LOGITS

Boo

0.78

ppo

0.76

 wedge

0.74

tein

0.71

ppel

0.70

 hect

0.70

wealth

0.68

ãĥĥãĤ¯

0.68

ĸļ

0.67

Activations Density 0.044%

words related to taxation and divisive or separating elements

language related to divisive issues or taxes

No Comments

No Known Activations