INDEX

Explanations

references to "articles" or "chronicles"

oai_token-act-pair · gpt-3.5-turbo

terms related to articles or headings in a document

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

itude

-0.93

itcher

-0.90

itches

-0.79

awaru

-0.78

atcher

-0.74

reath

-0.74

ļéĨĴ

-0.72

ublic

-0.71

gression

-0.71

raints

-0.71

POSITIVE LOGITS

icle

1.04

icles

0.97

oidal

0.74

ously

0.67

oids

0.66

phal

0.65

ect

0.64

 Britann

0.64

hog

0.64

 emit

0.64

Activations Density 0.033%

references to "articles" or "chronicles"

terms related to articles or headings in a document

No Comments

No Known Activations