INDEX

Explanations

technical instructions or commands

oai_token-act-pair · gpt-3.5-turbo

concepts related to health and safety

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

laughter

-0.80

___

-0.77

Newsletter

-0.77

TRUMP

-0.74

billion

-0.72

People

-0.71

Americans

-0.70

Everybody

-0.68

America

-0.66

CNN

-0.63

POSITIVE LOGITS

 util

0.91

 similarly

0.88

 whilst

0.86

 requisite

0.84

 aforementioned

0.84

 marginally

0.82

 subsequently

0.81

 comparatively

0.81

 Whilst

0.80

 extant

0.80

Activations Density 1.565%

technical instructions or commands

concepts related to health and safety

No Comments

No Known Activations