INDEX

Explanations

words related to incorrectness or errors

oai_token-act-pair · gpt-3.5-turbo

references to the concept of being incorrect or unjust

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hens

-0.76

enance

-0.70

 composed

-0.63

 Flavoring

-0.63

 Pione

-0.62

 Swim

-0.61

Fn

-0.60

kamp

-0.59

 Crunch

-0.59

hips

-0.59

POSITIVE LOGITS

headed

1.08

fully

1.07

 unfocusedRange

0.86

sight

0.85

eous

0.83

behavior

0.81

ftime

0.81

 assumptions

0.80

 guiActiveUn

0.80

do

0.79

Activations Density 0.032%

words related to incorrectness or errors

references to the concept of being incorrect or unjust

No Comments

No Known Activations