INDEX

Explanations

phrases related to justifying a particular action or decision

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate justification or rationale

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

mor

-0.64

Jar

-0.61

Dro

-0.60

las

-0.57

 mascara

-0.55

 Hunts

-0.55

thin

-0.55

 tolerated

-0.55

cause

-0.55

 choir

-0.55

POSITIVE LOGITS

SPONSORED

0.92

>>>>>>>>

0.70

nings

0.70

isphere

0.69

asion

0.68

achy

0.68

irth

0.68

akedown

0.67

gha

0.66

ativity

0.65

Activations Density 0.112%

phrases related to justifying a particular action or decision

phrases that indicate justification or rationale

No Comments

No Known Activations