INDEX

Explanations

references to watching, streaming, and downloading various forms of media like movies, TV shows, and music

oai_token-act-pair · gpt-3.5-turbo

references to media consumption and streaming services

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

peror

-0.73

 councill

-0.69

JM

-0.68

ahon

-0.67

phas

-0.66

resent

-0.64

wed

-0.63

rian

-0.63

 Cynthia

-0.63

stood

-0.62

POSITIVE LOGITS

netflix

0.86

 Mehran

0.83

Torrent

0.78

metadata

0.78

 msec

0.76

Netflix

0.73

clips

0.72

 Codec

0.72

Upload

0.70

 Mbps

0.69

Activations Density 0.310%

references to watching, streaming, and downloading various forms of media like movies, TV shows, and music

references to media consumption and streaming services

No Comments

No Known Activations