INDEX

Explanations

occurrences of the word "We" followed by varying contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the pronoun "We."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

quo

-0.75

 gratification

-0.75

LSD

-0.74

cum

-0.64

 reinforcement

-0.61

 PUBLIC

-0.60

UD

-0.59

 srfAttach

-0.58

 guiActiveUnfocused

-0.57

 rival

-0.57

POSITIVE LOGITS

bsite

1.08

ldon

1.08

bley

1.01

ighed

0.99

've

0.99

akening

0.98

alth

0.98

chwitz

0.98

're

0.97

asel

0.96

Activations Density 0.125%

occurrences of the word "We" followed by varying contexts

instances of the pronoun "We."

No Comments

No Known Activations