INDEX

Explanations

references to specific TV shows or characters from TV shows

oai_token-act-pair · gpt-3.5-turbo

references to the TV show "Arrow" and its related characters

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uration

-0.87

urers

-0.83

orget

-0.82

employment

-0.80

milo

-0.76

mble

-0.76

xual

-0.76

employed

-0.73

ured

-0.73

enance

-0.71

POSITIVE LOGITS

 Arrow

1.55

 Lantern

0.92

 Canary

0.81

 Canyon

0.74

Edge

0.74

 Creek

0.70

 Tool

0.68

 Dash

0.67

 Protocol

0.67

 Dexter

0.66

Activations Density 0.017%

references to specific TV shows or characters from TV shows

references to the TV show "Arrow" and its related characters

No Comments

No Known Activations