INDEX

Explanations

proper nouns, specifically names of products, services, and companies

oai_token-act-pair · gpt-3.5-turbo

instances of the word "called" 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 @ 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

IFrame

Link

Negative Logits

 istg

-0.86

olate

-0.78

bilt

-0.75

midt

-0.74

EEE

-0.73

=-=-

-0.72

REF

-0.71

olitics

-0.71

 guiName

-0.70

olitan

-0.69

POSITIVE LOGITS

calling

0.87

call

0.84

called

0.72

 attention

0.71

 Ares

0.71

 called

0.68

 Call

0.64

igraph

0.64

sighted

0.63

 namely

0.62

Activations Density 0.046%

proper nouns, specifically names of products, services, and companies

instances of the word "called" in various contexts

No Known Activations