INDEX

Explanations

mentions of the abbreviation "CE" followed by a number

oai_token-act-pair · gpt-3.5-turbo

references to CE (Common Era) dates and related terminology

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

nesota

-0.90

folk

-0.88

mort

-0.85

manager

-0.76

eness

-0.75

ificent

-0.74

fired

-0.74

owsky

-0.73

ogyn

-0.73

ãĥĦ

-0.72

POSITIVE LOGITS

VO

1.10

utical

0.99

lements

0.96

IRO

0.88

VERS

0.88

URRENT

0.88

LL

0.80

ASE

0.80

VER

0.80

LECT

0.77

Activations Density 0.019%

mentions of the abbreviation "CE" followed by a number

references to CE (Common Era) dates and related terminology

No Comments

No Known Activations