INDEX

Explanations

the word 'Roger'

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Roger."

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

ipeg

-0.75

runtime

-0.74

pron

-0.73

tnc

-0.71

morph

-0.68

erala

-0.67

spir

-0.67

ihar

-0.66

advert

-0.65

indu

-0.65

POSITIVE LOGITS

 Roger

1.23

Roger

1.17

otiation

0.85

 Goodell

0.83

rers

0.82

Rod

0.79

Wim

0.77

 Pens

0.76

 Waters

0.76

 Godd

0.74

Activations Density 0.006%

the word 'Roger'

mentions of the name "Roger."

No Comments

No Known Activations