INDEX

Explanations

information related to individuals including their names, activities, and accomplishments

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, particularly commas

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

ore

-0.67

ãĥ¥

-0.66

obin

-0.62

zing

-0.60

oir

-0.59

°

-0.58

orer

-0.56

rome

-0.56

ãĥ¼

-0.52

rage

-0.52

POSITIVE LOGITS

}"

0.90

 respectively

0.73

 denies

0.73

 flourished

0.71

 acknowledges

0.71

 drew

0.70

 congratulated

0.68

 namely

0.67

 prosec

0.67

 violates

0.66

Activations Density 0.538%

information related to individuals including their names, activities, and accomplishments

punctuation marks, particularly commas

No Comments

No Known Activations