INDEX

Explanations

numerical values in the context of measurements or rankings

oai_token-act-pair · gpt-3.5-turbo

numerical values or statistics

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

arial

-0.88

boarding

-0.77

mast

-0.74

antha

-0.73

many

-0.72

moon

-0.71

warts

-0.70

istically

-0.68

board

-0.66

igans

-0.65

POSITIVE LOGITS

é¾

0.77

uador

0.77

lectic

0.75

stasy

0.74

å¥³

0.72

TION

0.72

borg

0.69

æ©Ł

0.67

SHIP

0.64

Fuk

0.64

Activations Density 0.042%

numerical values in the context of measurements or rankings

numerical values or statistics

No Comments

No Known Activations