INDEX

Explanations

phrases related to opinions or stances on various issues

oai_token-act-pair · gpt-3.5-turbo

references to interpretations of moral codes and opinions about societal issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

Synopsis

-0.69

 Called

-0.64

xtap

-0.64

word

-0.64

cour

-0.60

Adds

-0.60

Spoiler

-0.58

Whe

-0.58

Incre

-0.58

Upon

-0.58

POSITIVE LOGITS

merce

0.72

ļé

0.70

 oneself

0.69

allery

0.67

actual

0.64

 Fiat

0.62

gans

0.62

ohm

0.60

 actual

0.60

 others

0.60

Activations Density 0.772%

phrases related to opinions or stances on various issues

references to interpretations of moral codes and opinions about societal issues

No Comments

No Known Activations