INDEX

Explanations

phrases indicating the context or setting in which something is happening

oai_token-act-pair · gpt-3.5-turbo

phrases and concepts that relate to context and perspective in discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

anooga

-0.80

unte

-0.75

uala

-0.75

alan

-0.73

ade

-0.67

ahs

-0.65

orie

-0.64

keys

-0.64

Fix

-0.62

ãĥį

-0.62

POSITIVE LOGITS

 context

0.96

 contexts

0.93

 broader

0.86

 contemporary

0.86

 socio

0.83

 totality

0.82

 everyday

0.80

xual

0.79

 realities

0.78

 discussing

0.78

Activations Density 0.331%

phrases indicating the context or setting in which something is happening

phrases and concepts that relate to context and perspective in discussions

No Comments

No Known Activations