INDEX

Explanations

phrases expressing opinions or emphasis

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "I mean" or similar expressions indicating clarification or emphasis

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

SourceFile

-0.84

MpServer

-0.80

Þ

-0.70

Recommended

-0.70

pec

-0.69

cloneembedreportprint

-0.68

ANI

-0.68

âĵĺ

-0.68

ÑĮ

-0.68

iHUD

-0.68

POSITIVE LOGITS

why

1.15

 imagine

1.04

hey

1.02

 yeah

1.00

 wouldn

0.98

 nobody

0.98

 suppose

0.90

if

0.87

 maybe

0.87

isn

0.87

Activations Density 0.170%

phrases expressing opinions or emphasis

instances of the phrase "I mean" or similar expressions indicating clarification or emphasis

No Comments

No Known Activations