INDEX

Explanations

phrases related to technical actions or operations

oai_token-act-pair · gpt-3.5-turbo

actions related to physical movement or tasks

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

 notor

-0.71

vae

-0.69

ometimes

-0.65

aples

-0.64

 Aires

-0.64

ĪĴ

-0.64

ayn

-0.63

 Argent

-0.62

 trave

-0.62

eger

-0.61

POSITIVE LOGITS

able

1.51

ings

1.36

ability

1.34

ables

1.19

athon

1.12

ers

1.10

away

1.05

outs

1.05

nings

1.03

aways

0.99

Activations Density 0.430%

phrases related to technical actions or operations

actions related to physical movement or tasks

No Comments

No Known Activations