INDEX

Explanations

references to the television network HBO

oai_token-act-pair · gpt-3.5-turbo

references to HBO and its shows

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uation

-0.86

acle

-0.86

omaly

-0.85

acular

-0.82

alion

-0.77

ional

-0.75

jury

-0.73

ertodd

-0.73

uart

-0.71

folk

-0.70

POSITIVE LOGITS

TV

0.89

HBO

0.84

 Films

0.83

EW

0.81

 Wire

0.80

IGHTS

0.77

ÃįÃį

0.74

ILLE

0.73

 Showtime

0.73

 Networks

0.73

Activations Density 0.025%

references to the television network HBO

references to HBO and its shows

No Comments

No Known Activations