INDEX

Explanations

mentions of individuals' names or specific entities

oai_token-act-pair · gpt-3.5-turbo

mentions of "ips" or phrases that refer to various contexts involving "ips."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Aval

-0.81

mates

-0.73

 Aura

-0.65

ãģ®å®

-0.65

 damned

-0.64

 Aqua

-0.63

 Saud

-0.62

WAY

-0.62

 beans

-0.61

 guiName

-0.61

POSITIVE LOGITS

ips

1.11

ilver

1.01

ilateral

0.99

etsk

0.99

osit

0.99

icket

0.97

aret

0.93

olitan

0.91

ongyang

0.90

ippi

0.90

Activations Density 0.007%

mentions of individuals' names or specific entities

mentions of "ips" or phrases that refer to various contexts involving "ips."

No Comments

No Known Activations