INDEX

Explanations

words related to skepticism or doubt

oai_token-act-pair · gpt-3.5-turbo

terms related to negative assessments or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

anwhile

-0.92

 Rover

-0.83

hyde

-0.72

 Agency

-0.70

 Agent

-0.68

MQ

-0.68

 Steps

-0.68

 Actions

-0.68

 Observatory

-0.67

 Tags

-0.67

POSITIVE LOGITS

structed

1.08

itial

0.99

stable

0.96

ceivable

0.95

plain

0.94

ufact

0.92

orable

0.91

verted

0.91

quer

0.89

vention

0.88

Activations Density 0.011%

words related to skepticism or doubt

terms related to negative assessments or conditions

No Comments

No Known Activations