INDEX

Explanations

phrases indicating some level of comparison or degree

oai_token-act-pair · gpt-3.5-turbo

parentheses used in sentences

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

arus

-0.75

oun

-0.71

monton

-0.69

akov

-0.68

oda

-0.67

icter

-0.66

resy

-0.65

APTER

-0.64

ibel

-0.64

igers

-0.62

POSITIVE LOGITS

 Malf

0.72

interest

0.69

multiple

0.62

suff

0.59

ãĥİ

0.59

hearted

0.57

 apologies

0.56

 Leilan

0.56

 inaction

0.55

stellar

0.54

Activations Density 0.089%

phrases indicating some level of comparison or degree

parentheses used in sentences

No Comments

No Known Activations