INDEX

Explanations

expressions related to numerical measurements of accuracy

oai_token-act-pair · gpt-3.5-turbo

references to negative values or measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

lied

-0.83

istle

-0.83

kamp

-0.77

uay

-0.74

akings

-0.73

ollen

-0.73

urrent

-0.72

bris

-0.70

anca

-0.70

stakes

-0.69

POSITIVE LOGITS

minus

1.07

cules

0.98

 minus

0.86

(-

0.72

/-

0.71

MSG

0.65

(-

0.64

 pinch

0.62

 heels

0.62

ï¸

0.62

Activations Density 0.004%

expressions related to numerical measurements of accuracy

references to negative values or measurements

No Comments

No Known Activations