INDEX

Explanations

phrases associated with instructions or commands

oai_token-act-pair · gpt-3.5-turbo

words related to advertising and promotion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

etheless

-0.62

 Frozen

-0.61

rawdownloadcloneembedreportprint

-0.61

 Guilty

-0.60

wra

-0.60

 Harbour

-0.59

Bug

-0.59

Chel

-0.58

 Newsp

-0.57

FN

-0.57

POSITIVE LOGITS

 Âł Âł

0.66

xon

0.66

Í

0.65

apeshifter

0.61

 transitions

0.60

 stride

0.59

 inclusive

0.58

 reconstruction

0.58

 gradient

0.58

)]

0.58

Activations Density 0.249%

phrases associated with instructions or commands

words related to advertising and promotion

No Comments

No Known Activations