INDEX

Explanations

words related to temporal distance, such as terms indicating approaching or approached

oai_token-act-pair · gpt-3.5-turbo

the word "approach" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

pite

-0.92

cell

-0.83

cake

-0.81

glass

-0.72

arus

-0.72

gans

-0.71

band

-0.69

gae

-0.68

fre

-0.65

uid

-0.65

POSITIVE LOGITS

 Ceres

0.88

 solicitation

0.68

 Lans

0.66

 heights

0.64

heny

0.62

 Pose

0.62

WARD

0.62

 thresholds

0.60

anasia

0.60

 nearer

0.60

Activations Density 0.025%

words related to temporal distance, such as terms indicating approaching or approached

the word "approach" and its variations

No Comments

No Known Activations