INDEX

Explanations

mentions of TV show seasons and episodes, as well as production details

oai_token-act-pair · gpt-3.5-turbo

references to television shows and their production details

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

jri

-0.67

 administr

-0.67

 confer

-0.64

Federal

-0.64

lyak

-0.60

 advant

-0.60

 routed

-0.60

 administ

-0.60

uration

-0.59

 exerc

-0.59

POSITIVE LOGITS

 sequels

1.29

 films

1.22

 novels

1.19

 movies

1.17

 screenplay

1.16

 thriller

1.16

 sitcom

1.14

 comedy

1.13

 trilogy

1.12

 sequel

1.11

Activations Density 1.000%

mentions of TV show seasons and episodes, as well as production details

references to television shows and their production details

No Comments

No Known Activations