INDEX

Explanations

phrases related to reasons or justifications

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize the importance of justification or reason behind actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 primed

-0.67

 intimidating

-0.63

bor

-0.61

 bron

-0.59

 Feld

-0.59

 penetration

-0.58

 infiltrated

-0.58

omer

-0.58

 Torn

-0.58

 feathers

-0.58

POSITIVE LOGITS

 sake

1.01

olicy

0.85

ames

0.83

ume

0.83

rior

0.79

atsu

0.79

uin

0.78

urities

0.77

¬¼

0.77

lihood

0.76

Activations Density 0.012%

phrases related to reasons or justifications

phrases that emphasize the importance of justification or reason behind actions

No Comments

No Known Activations