INDEX

Explanations

descriptions and mentions of various story elements

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the word "story" and related discussions about narrative elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

inence

-0.90

orem

-0.81

aurus

-0.75

ertodd

-0.73

agles

-0.70

usra

-0.69

oute

-0.68

heed

-0.67

olen

-0.66

ormonal

-0.65

POSITIVE LOGITS

telling

1.24

 arcs

1.15

arc

1.08

board

1.05

boards

1.05

 synopsis

0.97

nexus

0.89

book

0.88

boarding

0.88

line

0.86

Activations Density 0.065%

descriptions and mentions of various story elements

repeated mentions of the word "story" and related discussions about narrative elements

No Comments

No Known Activations