INDEX

Explanations

names or words related to "rose"

oai_token-act-pair · gpt-3.5-turbo

references to specific names or titles in media

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 Seym

-0.81

ilon

-0.78

ounced

-0.78

icably

-0.77

sonian

-0.75

 Hilbert

-0.70

orius

-0.70

chart

-0.70

odynamic

-0.67

adel

-0.67

POSITIVE LOGITS

xual

1.32

REDACTED

0.84

OUP

0.78

>>>>>>>>

0.76

phrine

0.74

eyes

0.68

isure

0.68

Movie

0.68

 Grove

0.67

Bron

0.67

Activations Density 0.033%

names or words related to "rose"

references to specific names or titles in media

No Known Activations