INDEX

Explanations

phrases related to measurement or assessment

oai_token-act-pair · gpt-3.5-turbo

expressions of scale and manner related to actions or conditions

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

 Doodle

-0.72

 Bridges

-0.68

 Branch

-0.66

blems

-0.63

 Rebell

-0.62

Dyn

-0.62

iffs

-0.61

Pap

-0.60

avorite

-0.59

 images

-0.59

POSITIVE LOGITS

adobe

0.79

MAL

0.74

aver

0.69

usually

0.68

 unspecified

0.66

ranging

0.66

fuck

0.65

cko

0.64

 unimaginable

0.64

 resembling

0.64

Activations Density 0.164%

phrases related to measurement or assessment

expressions of scale and manner related to actions or conditions

No Comments

No Known Activations