INDEX

Explanations

numeric values indicating a frequency or rate

oai_token-act-pair · gpt-3.5-turbo

instances of the word "each."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

etics

-0.67

Kap

-0.63

Cu

-0.60

env

-0.59

cards

-0.59

 Pact

-0.59

stones

-0.58

 Roberts

-0.57

 Journals

-0.57

IJ

-0.57

POSITIVE LOGITS

 month

1.23

 year

1.21

 week

1.14

 weekday

1.14

day

1.04

 semester

0.98

 successive

0.93

 morning

0.91

month

0.90

 Month

0.90

Activations Density 0.033%

numeric values indicating a frequency or rate

instances of the word "each."

No Comments

No Known Activations