INDEX

Explanations

the word "one" followed by a number indicating significance or importance

oai_token-act-pair · gpt-3.5-turbo

the term "one" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

osponsors

-0.74

ooks

-0.68

inders

-0.67

 actionGroup

-0.65

Available

-0.62

ãĥĥ

-0.57

 cliffs

-0.57

inas

-0.55

untu

-0.55

OAD

-0.55

POSITIVE LOGITS

 Hundred

0.95

 hundred

0.90

 Thousand

0.86

 thing

0.82

 thousand

0.77

rency

0.75

 dimensional

0.73

 sided

0.72

anchester

0.71

 Million

0.70

Activations Density 0.147%

the word "one" followed by a number indicating significance or importance

the term "one" in various contexts

No Comments

No Known Activations