INDEX

Explanations

instances of the word "return"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "return"

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

inx

-0.71

pees

-0.69

Cola

-0.63

forum

-0.62

ancies

-0.61

rum

-0.61

enges

-0.61

rote

-0.61

ãĥ»

-0.61

outed

-0.60

POSITIVE LOGITS

 return

3.73

return

2.62

 Return

2.51

 returns

2.49

Return

2.08

 returned

2.06

 returning

2.01

 Returns

1.82

 Returning

1.62

Returns

1.53

Activations Density 0.021%

instances of the word "return"

instances of the word "return"

No Comments

No Known Activations