INDEX

Explanations

names or titles related to leadership positions, particularly the title "Director."

oai_token-act-pair · gpt-3.5-turbo

references to individuals in directorial positions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asions

-0.66

theless

-0.65

 msec

-0.61

 mileage

-0.61

-0.59

adesh

-0.59

ulz

-0.59

 conver

-0.59

 Fighters

-0.58

Skill

-0.58

POSITIVE LOGITS

ial

1.06

ially

1.02

ovie

0.94

ials

0.94

IAL

0.85

director

0.85

ate

0.82

ates

0.82

itatively

0.80

Director

0.80

Activations Density 0.033%

names or titles related to leadership positions, particularly the title "Director."

references to individuals in directorial positions

No Comments

No Known Activations