INDEX

Explanations

words related to military operations and warfare

oai_token-act-pair · gpt-3.5-turbo

instances of the word "err" indicating errors or mistakes in context

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

esville

-0.77

xual

-0.72

eph

-0.69

ĺħ

-0.69

waves

-0.68

¥

-0.66

creen

-0.66

clud

-0.64

Sing

-0.64

rooms

-0.62

POSITIVE LOGITS

idge

0.98

rr

0.97

andom

0.95

ange

0.92

untled

0.91

antly

0.90

ands

0.87

utherford

0.87

anger

0.86

ific

0.84

Activations Density 0.016%

words related to military operations and warfare

instances of the word "err" indicating errors or mistakes in context

No Comments

No Known Activations