INDEX

Explanations

proper nouns related to characters and locations in a specific context, potentially from a movie or show

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and titles related to people and places

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Clement

-0.93

song

-0.87

Song

-0.77

ynes

-0.73

============

-0.72

SY

-0.71

icy

-0.71

 princess

-0.69

Prin

-0.69

cffffcc

-0.69

POSITIVE LOGITS

Kov

3.00

Ish

1.76

 Grind

1.58

 Kessler

1.43

 Grill

1.37

Sz

1.24

 grind

1.21

 Nish

1.19

 Slovakia

1.19

 Hutch

1.13

Activations Density 0.060%

proper nouns related to characters and locations in a specific context, potentially from a movie or show

proper nouns, particularly names and titles related to people and places

No Comments

No Known Activations