INDEX

Explanations

comparative words, particularly indicating a higher or lower quantity or degree

oai_token-act-pair · gpt-3.5-turbo

expressions related to increasing quantities or comparisons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Thomson

-0.74

 Films

-0.70

 Immun

-0.69

 Frankenstein

-0.68

Detroit

-0.66

 Zimmer

-0.65

xtap

-0.65

 Ferguson

-0.65

 Commissioner

-0.64

 Springer

-0.63

POSITIVE LOGITS

 than

1.26

 expensive

1.06

 stringent

0.94

 importantly

0.94

 manageable

0.92

ado

0.92

 complicated

0.92

natureconservancy

0.89

 efficient

0.89

 costly

0.88

Activations Density 0.160%

comparative words, particularly indicating a higher or lower quantity or degree

expressions related to increasing quantities or comparisons

No Comments

No Known Activations