INDEX

Explanations

words related to actions like pranks, providing, priding, or praying

oai_token-act-pair · gpt-3.5-turbo

words or phrases related to public relations or promotional activities

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

 Sacrament

-0.76

 Winds

-0.73

 Civil

-0.66

 Americas

-0.65

 breeze

-0.65

INESS

-0.65

 Cheong

-0.64

 Tome

-0.63

hiro

-0.62

dayName

-0.61

POSITIVE LOGITS

arians

1.14

imes

1.13

imate

1.12

uning

1.09

atical

1.05

icer

1.04

udence

1.03

icy

1.00

acking

1.00

ape

0.98

Activations Density 0.008%

words related to actions like pranks, providing, priding, or praying

words or phrases related to public relations or promotional activities

No Comments

No Known Activations