INDEX

Explanations

expressions related to quantitative measurements or statistical information, particularly focusing on percentages and margins

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate proportions or measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

BILL

-0.71

priv

-0.70

children

-0.70

pless

-0.68

roots

-0.66

ãĤ°

-0.66

è¦ļéĨĴ

-0.65

Downloadha

-0.65

é¾įå

-0.65

hunter

-0.65

POSITIVE LOGITS

 occurrence

0.96

 residence

0.90

 ignition

0.86

 engagement

0.86

 departure

0.84

 descent

0.81

 propulsion

0.81

 entry

0.81

 operation

0.80

rontal

0.80

Activations Density 0.118%

expressions related to quantitative measurements or statistical information, particularly focusing on percentages and margins

phrases that indicate proportions or measurements

No Comments

No Known Activations