INDEX

Explanations

references to various abbreviations containing the letters "IN."

oai_token-act-pair · gpt-3.5-turbo

instances of the term "IN" and its variations, indicating a focus on inclusion or involvement

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

apy

-0.68

 hatch

-0.60

 lands

-0.59

 resc

-0.58

 Kinder

-0.57

hood

-0.56

 bund

-0.56

 amph

-0.56

thouse

-0.56

gamer

-0.55

POSITIVE LOGITS

IN

3.77

INS

2.51

INE

2.15

INO

1.99

INC

1.87

INA

1.85

INT

1.77

INST

1.74

INAL

1.64

OUT

1.62

Activations Density 0.010%

references to various abbreviations containing the letters "IN."

instances of the term "IN" and its variations, indicating a focus on inclusion or involvement

No Comments

No Known Activations