INDEX

Explanations

numerical estimates or approximations

oai_token-act-pair · gpt-3.5-turbo

terms relating to estimates and approximations

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

NetMessage

-0.89

Interstitial

-0.84

woods

-0.78

feeding

-0.77

vous

-0.76

spr

-0.75

nut

-0.74

unin

-0.71

nuts

-0.70

kish

-0.67

POSITIVE LOGITS

 estimate

1.35

 estimates

1.26

imate

0.99

 estimation

0.98

imates

0.97

 estimating

0.97

 estimated

0.95

 Estimates

0.91

 estim

0.87

 allowance

0.80

Activations Density 0.014%

numerical estimates or approximations

terms relating to estimates and approximations

No Comments

No Known Activations