INDEX

Explanations

phrases expressing similarity or continuity

oai_token-act-pair · gpt-3.5-turbo

statements asserting the validity or truth of a proposition across various contexts

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

 helicop

-0.80

 deserts

-0.70

casters

-0.67

 calcul

-0.67

 escal

-0.61

 flares

-0.61

inker

-0.60

 ramps

-0.60

TAG

-0.59

LIM

-0.59

POSITIVE LOGITS

kefeller

0.79

VB

0.79

same

0.77

ItemThumbnailImage

0.75

 Tradable

0.74

oother

0.74

":"/

0.74

 Same

0.72

^^^^

0.72

 Berm

0.71

Activations Density 0.211%

phrases expressing similarity or continuity

statements asserting the validity or truth of a proposition across various contexts

No Comments

No Known Activations