INDEX

Explanations

phrases related to causes or reasons

oai_token-act-pair · gpt-3.5-turbo

references to various causes and their implications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

aeper

-0.81

PDATE

-0.77

 Pione

-0.75

 Leopard

-0.70

 Technique

-0.69

ault

-0.68

esters

-0.67

 Sheep

-0.65

McM

-0.63

town

-0.63

POSITIVE LOGITS

 cele

1.14

 Causes

0.87

 causes

0.84

 Cause

0.83

 cause

0.82

cause

0.81

hift

0.79

 celeb

0.75

hooting

0.70

wagon

0.69

Activations Density 0.007%

phrases related to causes or reasons

references to various causes and their implications

No Comments

No Known Activations