INDEX

Explanations

phrases related to ratings

oai_token-act-pair · gpt-3.5-turbo

mentions of ratings or evaluations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gnu

-0.74

pora

-0.72

Alz

-0.71

ensis

-0.68

ctuary

-0.68

 ancestor

-0.67

ford

-0.67

vas

-0.67

 happ

-0.66

plain

-0.66

POSITIVE LOGITS

 ratings

3.98

 Ratings

3.06

 rating

2.64

 Rating

2.11

Rating

1.89

 rated

1.74

rating

1.67

 rankings

1.62

 viewership

1.61

 reviews

1.47

Activations Density 0.008%

phrases related to ratings

mentions of ratings or evaluations

No Comments

No Known Activations