INDEX

Explanations

the word "is" at a high activation level

oai_token-act-pair · gpt-3.5-turbo

instances of the verb "is."

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

*/

-0.63

 entails

-0.63

 refrain

-0.61

ependence

-0.58

know

-0.58

 encounters

-0.57

urry

-0.57

 wills

-0.57

itiveness

-0.56

 IMAGES

-0.56

POSITIVE LOGITS

senal

1.05

rael

1.00

cussion

0.90

 definitely

0.88

 rumored

0.85

 currently

0.83

nt

0.82

 hereby

0.82

 indeed

0.81

 supposed

0.81

Activations Density 0.408%

the word "is" at a high activation level

instances of the verb "is."

No Comments

No Known Activations