INDEX

Explanations

number values in a specific range

oai_token-act-pair · gpt-3.5-turbo

numerical values, particularly those associated with thresholds or measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ItemImage

-0.65

ercise

-0.63

akov

-0.63

 indo

-0.59

nice

-0.58

perty

-0.56

 Spears

-0.56

 cold

-0.55

 refrain

-0.53

tsky

-0.53

POSITIVE LOGITS

xff

1.35

xa

1.08

xd

1.08

xc

1.03

xb

1.01

0.92

0.91

xe

0.91

0.89

Activations Density 0.033%

number values in a specific range

numerical values, particularly those associated with thresholds or measurements

No Comments

No Known Activations