INDEX

Explanations

words related to changes or differences

oai_token-act-pair · gpt-3.5-turbo

mentions of change, particularly in a variety of 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

soever

-0.75

EEEE

-0.66

rs

-0.64

linger

-0.64

lynn

-0.63

WER

-0.60

Nat

-0.59

Wire

-0.59

OHN

-0.58

IDE

-0.58

POSITIVE LOGITS

effic

1.13

ordinate

1.12

 relation

1.11

efficiency

1.10

 regards

1.06

humane

1.03

between

1.02

 favor

1.02

animate

0.99

clusions

0.99

Activations Density 0.168%

words related to changes or differences

mentions of change, particularly in a variety of contexts

No Comments

No Known Activations