INDEX

Explanations

the names of actors starring in films

oai_token-act-pair · gpt-3.5-turbo

instances of the word "in" used in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

llor

-0.82

 incumb

-0.72

bryce

-0.69

soever

-0.68

akia

-0.68

 inconven

-0.67

 incl

-0.65

killed

-0.65

 defect

-0.65

ileaks

-0.64

POSITIVE LOGITS

 lieu

1.33

 conjunction

1.17

clus

1.16

 accordance

1.07

 spite

1.07

 regards

1.05

 unison

1.05

animate

1.02

 addition

1.00

 disguise

1.00

Activations Density 0.562%

the names of actors starring in films

instances of the word "in" used in various contexts

No Comments

No Known Activations