INDEX

Explanations

words related to initial actions or stages of events

oai_token-act-pair · gpt-3.5-turbo

the recurring use of the word "initially."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

rf

-0.77

masters

-0.77

 Vish

-0.70

 Chef

-0.68

cig

-0.67

rom

-0.67

=-=-=-=-=-=-=-=-

-0.67

laws

-0.65

gerald

-0.64

Cth

-0.64

POSITIVE LOGITS

 responders

0.89

 blush

0.87

 conceived

0.80

 hesitated

0.80

 unsuccessfully

0.78

 stages

0.78

 appeared

0.77

 solic

0.77

 plotted

0.76

 initialized

0.74

Activations Density 0.013%

words related to initial actions or stages of events

the recurring use of the word "initially."

No Comments

No Known Activations