INDEX

Explanations

words related to someone taking an action or making a decision

oai_token-act-pair · gpt-3.5-turbo

occurrences of the verb "took."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

eers

-0.74

Cong

-0.70

 david

-0.70

 ---------

-0.63

 Smile

-0.63

holm

-0.62

egg

-0.62

nep

-0.59

ieg

-0.59

iege

-0.57

POSITIVE LOGITS

 aback

0.95

aways

0.92

 advantage

0.89

ãĤ¤ãĥĪ

0.84

FINE

0.81

 heed

0.80

arnaev

0.78

autions

0.76

ume

0.75

weed

0.74

Activations Density 0.096%

words related to someone taking an action or making a decision

occurrences of the verb "took."

No Comments

No Known Activations