INDEX

Explanations

words related to viewpoint or interpretation

oai_token-act-pair · gpt-3.5-turbo

references to varying viewpoints or perceptions

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

avis

-0.77

liam

-0.74

nard

-0.68

nar

-0.68

cakes

-0.68

die

-0.67

ertodd

-0.66

icit

-0.65

Interstitial

-0.65

ldon

-0.64

POSITIVE LOGITS

 perspectives

0.96

 perspective

0.92

 viewpoint

0.88

 viewpoints

0.83

 Perspective

0.78

 views

0.76

spection

0.74

 lens

0.67

 Lens

0.67

yip

0.67

Activations Density 0.021%

words related to viewpoint or interpretation

references to varying viewpoints or perceptions

No Comments

No Known Activations