INDEX

Explanations

the word "common"

oai_token-act-pair · gpt-3.5-turbo

references to the term "common" 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

oby

-0.72

ackers

-0.71

wards

-0.70

AA

-0.69

udging

-0.68

forestation

-0.67

hyde

-0.63

 fres

-0.62

haw

-0.62

iang

-0.62

POSITIVE LOGITS

wealth

1.14

alities

0.97

ality

0.91

 Common

0.89

 common

0.89

 occurrence

0.87

ensical

0.85

 occurrences

0.84

Common

0.81

 ancestor

0.78

Activations Density 0.015%

the word "common"

references to the term "common" in various contexts

No Comments

No Known Activations