INDEX

Explanations

mentions of the action verb "see"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "see."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

EStreamFrame

-0.79

cult

-0.73

 disposal

-0.70

prime

-0.69

ortunately

-0.69

 whistle

-0.69

cÃ©

-0.68

locked

-0.66

trap

-0.64

cious

-0.63

POSITIVE LOGITS

enance

0.84

seeing

0.84

¿½

0.80

 whence

0.78

see

0.77

erala

0.75

lights

0.72

saw

0.71

ablishment

0.71

adow

0.70

Activations Density 0.062%

mentions of the action verb "see"

occurrences of the word "see."

No Comments

No Known Activations