INDEX

Explanations

calls to action related to signing up for a newsletter

oai_token-act-pair · gpt-3.5-turbo

calls to action relating to signing up or subscribing

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Manson

-0.74

>]

-0.70

bis

-0.70

ranch

-0.70

mong

-0.64

tal

-0.62

_>

-0.62

gran

-0.61

 Fidel

-0.61

POSITIVE LOGITS

grades

0.84

rights

0.81

grading

0.80

 WATCHED

0.78

olicy

0.78

Trend

0.77

dates

0.77

np

0.77

 Tradable

0.76

ername

0.76

Activations Density 0.019%

calls to action related to signing up for a newsletter

calls to action relating to signing up or subscribing

No Comments

No Known Activations