INDEX

Explanations

words related to things becoming popular or widespread on the internet or social media

oai_token-act-pair · gpt-3.5-turbo

terms related to public announcements and viral events

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

ggles

-0.72

anges

-0.62

tein

-0.61

section

-0.61

give

-0.60

pri

-0.59

MOR

-0.58

jection

-0.57

outer

-0.57

ilege

-0.57

POSITIVE LOGITS

DAQ

0.95

 havoc

0.84

ousel

0.74

 unnoticed

0.73

aneously

0.71

Reloaded

0.70

 throats

0.70

 onstage

0.67

ãĢĤ

0.66

 withd

0.65

Activations Density 0.080%

words related to things becoming popular or widespread on the internet or social media

terms related to public announcements and viral events

No Comments

No Known Activations