INDEX

Explanations

words related to military presence or control of territory

oai_token-act-pair · gpt-3.5-turbo

references to various forms of occupation

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

disc

-0.83

ectar

-0.74

peak

-0.72

enic

-0.71

tle

-0.69

lua

-0.69

orah

-0.66

ramid

-0.66

cgi

-0.66

culated

-0.65

POSITIVE LOGITS

 occupation

1.14

 occup

0.94

 occupations

0.93

eering

0.86

 Occupations

0.83

isans

0.81

ãĤ¢ãĥ«

0.80

 occupying

0.79

 warr

0.77

 Occup

0.76

Activations Density 0.005%

words related to military presence or control of territory

references to various forms of occupation

No Comments

No Known Activations