INDEX

Explanations

references to on-screen visuals or actions

oai_token-act-pair · gpt-3.5-turbo

references to "screen" 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

IGH

-0.77

ETA

-0.72

ghan

-0.69

 deaf

-0.65

 rush

-0.65

 punishing

-0.64

ATHER

-0.64

tta

-0.63

ggies

-0.62

 frig

-0.61

POSITIVE LOGITS

screen

1.21

 Screen

1.03

Screen

0.94

reens

0.94

TVs

0.91

wip

0.91

 screenings

0.88

 screens

0.83

 screen

0.81

osterone

0.79

Activations Density 0.013%

references to on-screen visuals or actions

references to "screen" in various contexts

No Comments

No Known Activations