INDEX

Explanations

proper nouns or specific names related to organizations, companies, or entities

oai_token-act-pair · gpt-3.5-turbo

references to named entities or proper nouns

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rio

-0.80

ingham

-0.77

avia

-0.74

borgh

-0.74

rums

-0.73

rian

-0.73

hips

-0.72

gian

-0.71

zona

-0.70

aunt

-0.70

POSITIVE LOGITS

NP

1.22

NP

1.12

oint

0.88

ointed

0.78

ointment

0.77

emonic

0.75

isites

0.73

hemer

0.72

 Advertisement

0.70

olicy

0.67

Activations Density 0.007%

proper nouns or specific names related to organizations, companies, or entities

references to named entities or proper nouns

No Comments

No Known Activations