INDEX

Explanations

numerical values with or without decimals

oai_token-act-pair · gpt-3.5-turbo

occurrences of numerical data or statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

urized

-0.79

umbn

-0.77

 outl

-0.68

 primaries

-0.67

 foreseeable

-0.64

 suprem

-0.63

 subsequ

-0.61

achus

-0.61

 accrued

-0.61

 consecut

-0.60

POSITIVE LOGITS

©¶æ¥µ

0.81

 ILCS

0.78

 Introduction

0.76

Hel

0.74

çīĪ

0.73

 Expand

0.68

 Inquisitor

0.68

MPG

0.68

 Detected

0.66

MIL

0.65

Activations Density 0.059%

numerical values with or without decimals

occurrences of numerical data or statistics

No Comments

No Known Activations