INDEX

Explanations

mentions of locations or entities with the prefix "Mo"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Mo" or related constructs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 responsibility

-0.71

 Icelandic

-0.69

 Identification

-0.68

 Integrity

-0.68

 Prohibition

-0.66

 Discrimination

-0.62

 caution

-0.62

 Prelude

-0.62

stood

-0.61

 envy

-0.60

POSITIVE LOGITS

ose

1.14

oney

1.06

oby

1.00

omon

0.97

ogly

0.93

osh

0.93

ascus

0.92

oser

0.92

utes

0.91

oses

0.90

Activations Density 0.012%

mentions of locations or entities with the prefix "Mo"

occurrences of the word "Mo" or related constructs

No Comments

No Known Activations