INDEX

Explanations

words related to government, politics, and power dynamics

oai_token-act-pair · gpt-3.5-turbo

strongly activated sequences or mentions of "str" that signify significant or strategic 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

¥µ

-0.69

 timely

-0.63

 smoke

-0.62

mits

-0.61

pages

-0.60

dry

-0.60

IAS

-0.60

delay

-0.59

wolves

-0.59

 writing

-0.59

POSITIVE LOGITS

ategy

1.08

ategic

1.02

uments

0.99

icken

0.98

str

0.97

aditional

0.96

angles

0.95

angled

0.94

ained

0.94

acket

0.93

Activations Density 0.006%

words related to government, politics, and power dynamics

strongly activated sequences or mentions of "str" that signify significant or strategic elements in various contexts

No Comments

No Known Activations