INDEX

Explanations

conditional sentences expressing hypothetical situations

oai_token-act-pair · gpt-3.5-turbo

conditional statements about potential outcomes

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

²¾

-0.81

Ŀ

-0.75

{:

-0.67

ãĥ¼ãĥĨãĤ£

-0.63

ãĤ¼

-0.62

ãĥ£

-0.61

 Moroc

-0.60

endish

-0.59

EV

-0.59

vable

-0.58

POSITIVE LOGITS

glers

0.66

then

0.65

aples

0.64

CHA

0.64

 Cooke

0.60

 then

0.59

'll

0.59

 THEN

0.58

 elig

0.58

 SHALL

0.58

Activations Density 0.493%

conditional sentences expressing hypothetical situations

conditional statements about potential outcomes

No Comments

No Known Activations