INDEX

Explanations

phrases related to the term "Hack"

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "Hack" and related hacking concepts

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

 compr

-0.79

 confir

-0.79

 eleph

-0.74

iatrics

-0.71

 occas

-0.70

 appar

-0.68

 assassinate

-0.68

 Demonic

-0.68

CPC

-0.67

åĬ

-0.66

POSITIVE LOGITS

aday

1.23

ers

1.16

enberg

1.11

intosh

1.08

ett

0.98

athon

0.96

ership

0.94

age

0.93

Hack

0.92

haus

0.91

Activations Density 0.032%

phrases related to the term "Hack"

mentions of the term "Hack" and related hacking concepts

No Comments

No Known Activations