INDEX

Explanations

text related to balance, whether it's a balanced view, explanation, or diet

oai_token-act-pair · gpt-3.5-turbo

expressions related to the concept of balance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

olog

-0.79

clips

-0.78

OLOG

-0.76

ABE

-0.73

ography

-0.72

ACH

-0.72

 Offic

-0.72

chief

-0.71

Nazi

-0.70

Tat

-0.69

POSITIVE LOGITS

balanced

1.28

 balanced

1.23

 imbalance

1.12

 Balanced

0.96

 balancing

0.96

 balance

0.92

balance

0.92

 parity

0.90

 balances

0.89

 neutrality

0.82

Activations Density 0.008%

text related to balance, whether it's a balanced view, explanation, or diet

expressions related to the concept of balance

No Comments

No Known Activations