INDEX

Explanations

phrases where something/someone is being compared as being equally something else

oai_token-act-pair · gpt-3.5-turbo

the concept of equality or equal treatment 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

 Canaver

-0.75

liest

-0.74

ASE

-0.67

 Origins

-0.67

scape

-0.65

ocene

-0.65

ger

-0.64

history

-0.64

Query

-0.63

glers

-0.62

POSITIVE LOGITS

 situated

0.92

 spaced

0.90

 minded

0.86

imilar

0.83

 equally

0.82

 sized

0.81

inguishable

0.78

 applicable

0.77

etrical

0.76

ministic

0.73

Activations Density 0.013%

phrases where something/someone is being compared as being equally something else

the concept of equality or equal treatment in various contexts

No Comments

No Known Activations