INDEX

Explanations

media-related terms or references

oai_token-act-pair · gpt-3.5-turbo

references to media entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 turn

-0.78

 turns

-0.73

unts

-0.72

 bondage

-0.69

aph

-0.69

 goblin

-0.68

 squat

-0.66

 Pumpkin

-0.65

 squats

-0.65

 goblins

-0.65

POSITIVE LOGITS

Media

4.22

media

2.23

 Media

2.09

Image

1.79

 media

1.51

Video

1.40

Images

1.28

Press

1.26

imedia

1.25

MED

1.25

Activations Density 0.011%

media-related terms or references

references to media entities

No Comments

No Known Activations