INDEX

Explanations

phrases indicating hearsay or reported information

oai_token-act-pair · gpt-3.5-turbo

phrases indicating actions or states of being, particularly those introduced by "to be" and "to have."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Sorry

-0.67

noticed

-0.65

Letter

-0.65

APTER

-0.63

patch

-0.63

 realities

-0.63

 reminder

-0.63

leaf

-0.62

 Courier

-0.62

river

-0.60

POSITIVE LOGITS

 originate

0.92

 have

0.87

 embody

0.85

 derive

0.84

be

0.83

 indicate

0.79

 undergo

0.75

 anecd

0.75

 relate

0.74

 comprise

0.74

Activations Density 0.064%

phrases indicating hearsay or reported information

phrases indicating actions or states of being, particularly those introduced by "to be" and "to have."

No Comments

No Known Activations