INDEX

Explanations

phrases related to the term "IP."

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "nip" or "nipping."

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

 EntityItem

-0.87

DERR

-0.82

é¾

-0.72

 withd

-0.71

BIP

-0.67

 Baldwin

-0.64

 Cabrera

-0.63

DEV

-0.63

 Aval

-0.63

 Dying

-0.62

POSITIVE LOGITS

artisan

1.46

otent

1.30

osit

1.19

ublic

1.12

olitan

1.07

resent

1.06

olitics

1.05

rint

1.05

rison

1.05

inion

1.02

Activations Density 0.039%

phrases related to the term "IP."

mentions of the term "nip" or "nipping."

No Comments

No Known Activations