INDEX

Explanations

phrases related to actions causing change or consequences

oai_token-act-pair · gpt-3.5-turbo

actions related to transformation or change

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

ancies

-0.62

idence

-0.59

inia

-0.56

 occupancy

-0.56

redo

-0.54

rise

-0.54

dom

-0.54

awks

-0.53

ajo

-0.53

brate

-0.53

POSITIVE LOGITS

 hostage

0.74

aundering

0.73

Ń·

0.67

ety

0.67

../

0.66

 arbitrarily

0.66

UC

0.64

ĸļ

0.64

YING

0.64

by

0.64

Activations Density 0.226%

phrases related to actions causing change or consequences

actions related to transformation or change

No Comments

No Known Activations