INDEX

Explanations

phrases related to new information or discoveries

oai_token-act-pair · gpt-3.5-turbo

phrases related to the emergence or revelation of information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ingham

-0.70

aldi

-0.69

MSG

-0.67

 Editors

-0.65

 Princ

-0.64

gee

-0.64

mun

-0.59

 Auckland

-0.58

ongyang

-0.58

 Gord

-0.57

POSITIVE LOGITS

Film

0.80

heet

0.73

ixel

0.72

ymes

0.71

 spontaneously

0.70

Question

0.67

utra

0.67

IFF

0.66

EVA

0.65

quished

0.64

Activations Density 0.090%

phrases related to new information or discoveries

phrases related to the emergence or revelation of information

No Comments

No Known Activations