INDEX

Explanations

phrases related to public opinion or sentiment

oai_token-act-pair · gpt-3.5-turbo

terms related to public opinion and support

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

tein

-0.94

rete

-0.71

oret

-0.69

risome

-0.67

acter

-0.66

porter

-0.63

achu

-0.63

 Offic

-0.63

mers

-0.62

bial

-0.61

POSITIVE LOGITS

SPONSORED

0.95

 polls

0.77

 initiatives

0.74

 indexes

0.72

atures

0.72

 rating

0.72

MENTS

0.70

 surveys

0.70

 judgments

0.69

 favoring

0.67

Activations Density 0.227%

phrases related to public opinion or sentiment

terms related to public opinion and support

No Comments

No Known Activations