INDEX

Explanations

adjectives expressing superiority or inferiority

oai_token-act-pair · gpt-3.5-turbo

references to superiority and inferiority in various contexts

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

mberg

-0.93

BRE

-0.86

 Dresden

-0.85

gren

-0.75

RPG

-0.75

sol

-0.75

OWS

-0.75

der

-0.72

tw

-0.71

kj

-0.70

POSITIVE LOGITS

 dilig

0.91

intendent

0.90

xual

0.81

ority

0.78

itized

0.77

 grades

0.76

 superior

0.75

 bidder

0.75

 intellect

0.73

 nomine

0.73

Activations Density 0.017%

adjectives expressing superiority or inferiority

references to superiority and inferiority in various contexts

No Comments

No Known Activations