INDEX

Explanations

words containing the substring "wo"

oai_token-act-pair · gpt-3.5-turbo

words or terms that begin with "wo" or pertain to whimsical or fantastical themes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Emin

-0.79

 Spiral

-0.71

 Friedman

-0.70

aligned

-0.70

 Moreno

-0.68

 Gilmore

-0.68

 Showtime

-0.67

approved

-0.67

MAP

-0.66

enced

-0.66

POSITIVE LOGITS

wo

3.61

wo

2.65

Wo

2.23

Wo

2.03

 baff

1.19

 mammoth

1.15

twe

1.12

zo

0.97

oh

0.89

 ponder

0.88

Activations Density 0.027%

words containing the substring "wo"

words or terms that begin with "wo" or pertain to whimsical or fantastical themes

No Comments

No Known Activations