INDEX

Explanations

phrases related to step-by-step processes or instructions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating processes or methods that are carried out systematically

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

©¶æ¥µ

-0.72

ŃĶ

-0.59

 recy

-0.57

nu

-0.56

Nab

-0.56

Virgin

-0.56

NEC

-0.55

Emb

-0.53

 confessions

-0.53

 Nicotine

-0.52

POSITIVE LOGITS

dden

0.78

0.71

_-_

0.70

wards

0.70

ward

0.69

pherd

0.67

Ó

0.66

pping

0.65

verse

0.64

visory

0.63

Activations Density 0.022%

phrases related to step-by-step processes or instructions

phrases indicating processes or methods that are carried out systematically

No Comments

No Known Activations