INDEX

Explanations

words related to official documents, announcements, or writings like "whitepaper" or "depicted"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "ep," which may relate to specific topics or contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 referen

-0.82

 nomine

-0.80

ModLoader

-0.79

////////////////

-0.73

Ô

-0.73

 adolesc

-0.71

 glim

-0.71

ikuman

-0.71

================================================================

-0.70

hips

-0.69

POSITIVE LOGITS

isodes

1.27

ublic

1.17

odcast

1.16

stakes

1.13

artment

1.08

rint

1.07

iphany

1.04

hemer

1.04

edia

1.00

osit

0.97

Activations Density 0.013%

words related to official documents, announcements, or writings like "whit**ep**aper" or "d**ep**icted"

occurrences of the term "ep," which may relate to specific topics or contexts

No Comments

No Known Activations

words related to official documents, announcements, or writings like "whitepaper" or "depicted"