INDEX

Explanations

information related to TV show episodes, specifically the episode numbers

oai_token-act-pair · gpt-3.5-turbo

mentions of specific episodes from a series

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

ntil

-0.92

enei

-0.75

erness

-0.75

hari

-0.75

gars

-0.72

minist

-0.71

hirt

-0.71

umbai

-0.71

punk

-0.71

sheets

-0.70

POSITIVE LOGITS

 episode

0.92

 finale

0.92

 aired

0.87

 episodes

0.86

 airing

0.84

 Transcript

0.83

 opener

0.83

 premiere

0.81

 airs

0.79

 premie

0.79

Activations Density 0.031%

information related to TV show episodes, specifically the episode numbers

mentions of specific episodes from a series

No Comments

No Known Activations