INDEX

Explanations

words related to cause and effect

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "effect."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

idel

-0.78

oway

-0.65

 SATA

-0.63

lehem

-0.61

bledon

-0.60

gio

-0.59

dar

-0.59

itars

-0.59

ritz

-0.58

nia

-0.58

POSITIVE LOGITS

 effect

3.83

 Effect

2.67

 effects

2.40

effect

2.34

Effect

2.18

 Effects

1.97

effects

1.85

 impact

1.84

Effects

1.74

 affect

1.56

Activations Density 0.015%

words related to cause and effect

references to the concept of "effect."

No Comments

No Known Activations