INDEX

Explanations

mentions of individuals who were previously employed in notable positions

oai_token-act-pair · gpt-3.5-turbo

references to individuals with previous job titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

allery

-0.77

andise

-0.77

otle

-0.76

ammy

-0.75

anguage

-0.73

matic

-0.72

olen

-0.72

antics

-0.71

aris

-0.71

utterstock

-0.71

POSITIVE LOGITS

 Yugoslav

1.02

 Yugoslavia

0.92

 employee

0.82

 classmate

0.81

 pupil

0.81

 Soviet

0.78

foundland

0.77

 president

0.75

 vice

0.75

 journalist

0.73

Activations Density 0.029%

mentions of individuals who were previously employed in notable positions

references to individuals with previous job titles

No Comments

No Known Activations