INDEX

Explanations

phrases related to weaponry and military activities, or locations

oai_token-act-pair · gpt-3.5-turbo

terms associated with various types of physical objects and arrangements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

amaz

-0.71

ModLoader

-0.65

adelphia

-0.63

 mattress

-0.63

asus

-0.62

aeper

-0.62

Reloaded

-0.59

noon

-0.59

-0.58

ALT

-0.58

POSITIVE LOGITS

 alike

2.02

 respectively

1.47

 thereof

1.14

paces

1.12

hips

1.03

etc

0.92

hops

0.90

 abound

0.86

uits

0.85

 depending

0.84

Activations Density 0.329%

phrases related to weaponry and military activities, or locations

terms associated with various types of physical objects and arrangements

No Comments

No Known Activations