INDEX

Explanations

references to influential positions or prominent entities

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "vy" or related variants

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

Reviewed

-0.78

emate

-0.75

eared

-0.68

emark

-0.67

hetic

-0.66

aphael

-0.66

yrinth

-0.65

rador

-0.65

Contents

-0.64

Haunted

-0.63

POSITIVE LOGITS

vy

1.17

puff

0.85

yy

0.85

anka

0.83

eus

0.80

0.75

bags

0.73

hawks

0.69

plets

0.69

reys

0.69

Activations Density 0.025%

references to influential positions or prominent entities

occurrences of the word "vy" or related variants

No Comments

No Known Activations