INDEX

Explanations

instances of the word "not" in phrases indicating a negative outlook or obstacle

oai_token-act-pair · gpt-3.5-turbo

the negation of statements or propositions

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

ngth

-0.88

æ©

-0.66

actionDate

-0.63

Ô

-0.62

velength

-0.61

letter

-0.61

 Boxing

-0.60

lance

-0.60

itor

-0.59

ãĥ£

-0.59

POSITIVE LOGITS

epad

0.95

 necessarily

0.92

icably

0.83

eworthy

0.83

 technically

0.82

 exactly

0.81

 entirely

0.79

 outright

0.77

 intimidated

0.75

 altogether

0.74

Activations Density 0.030%

instances of the word "not" in phrases indicating a negative outlook or obstacle

the negation of statements or propositions

No Comments

No Known Activations