INDEX

Explanations

phrases related to measurements and names

oai_token-act-pair · gpt-3.5-turbo

names and measurements related to individuals and units of measurement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

userc

-0.89

icked

-0.78

execute

-0.74

 pestic

-0.74

gradient

-0.74

ids

-0.74

raltar

-0.73

 opio

-0.72

aukee

-0.72

andre

-0.71

POSITIVE LOGITS

 reprint

0.92

McA

0.78

 Carlson

0.76

 Berman

0.75

leen

0.74

 Ital

0.71

 Concord

0.67

 Worcester

0.67

 Reich

0.66

Wis

0.65

Activations Density 0.010%

phrases related to measurements and names

names and measurements related to individuals and units of measurement

No Comments

No Known Activations