INDEX

Explanations

words related to experience or state of being

oai_token-act-pair · gpt-3.5-turbo

instances of the word "being" 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 @ 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

inav

-0.78

 constellation

-0.73

PsyNetMessage

-0.73

 fray

-0.71

NetMessage

-0.63

 Strait

-0.62

orrow

-0.61

 Times

-0.61

 heavens

-0.61

 slope

-0.59

POSITIVE LOGITS

 able

1.29

 bitten

0.89

 judged

0.88

 aware

0.86

held

0.86

 unable

0.85

 notified

0.85

 eaten

0.84

 transported

0.80

 briefed

0.80

Activations Density 0.080%

words related to experience or state of being

instances of the word "being" in various contexts

No Comments

No Known Activations