INDEX

Explanations

titles or mentions related to the Star Wars franchise

oai_token-act-pair · gpt-3.5-turbo

references to the Star Wars franchise

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

rac

-0.88

ours

-0.88

bda

-0.85

hello

-0.85

absolute

-0.85

mble

-0.85

ent

-0.84

inner

-0.82

our

-0.81

hor

-0.80

POSITIVE LOGITS

 Rapids

0.85

 Islanders

0.83

 Wire

0.81

 Oaks

0.80

 Jindal

0.79

 Shield

0.79

 Mexicans

0.78

 Fury

0.77

 Lord

0.76

 Patriot

0.75

Activations Density 0.400%

titles or mentions related to the Star Wars franchise

references to the Star Wars franchise

No Comments

No Known Activations