INDEX

Explanations

adjectives related to negative characteristics

oai_token-act-pair · gpt-3.5-turbo

words indicating deceitful or harmful qualities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

PA

-0.64

 position

-0.63

peg

-0.62

 stand

-0.62

 life

-0.60

AMA

-0.60

 transcript

-0.60

 projected

-0.59

 termin

-0.58

AA

-0.58

POSITIVE LOGITS

icious

4.76

eous

1.08

volent

1.02

ious

0.99

oaded

0.99

itious

0.99

ulent

0.98

ices

0.97

icy

0.96

cious

0.96

Activations Density 0.013%

adjectives related to negative characteristics

words indicating deceitful or harmful qualities

No Comments

No Known Activations