INDEX

Explanations

specifically named items or concepts

oai_token-act-pair · gpt-3.5-turbo

the term "namely" and its related uses to introduce examples or specifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

-0.72

 Bulletin

-0.68

osc

-0.67

 Hunt

-0.66

remlin

-0.64

nee

-0.63

 Bagg

-0.63

obal

-0.63

age

-0.63

olog

-0.60

POSITIVE LOGITS

 namely

1.06

soever

0.85

ignty

0.79

viz

0.77

xual

0.74

atoon

0.73

ional

0.73

 ours

0.72

æ©Ł

0.71

ngth

0.71

Activations Density 0.005%

specifically named items or concepts

the term "namely" and its related uses to introduce examples or specifications

No Comments

No Known Activations