INDEX

Explanations

instructions or steps in a recipe

oai_token-act-pair · gpt-3.5-turbo

cooking actions and instructions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

DN

-0.67

anmar

-0.66

âĹ¼

-0.65

ãĤ´ãĥ³

-0.62

 successor

-0.62

iami

-0.62

furt

-0.61

SEE

-0.61

inventoryQuantity

-0.60

Earn

-0.59

POSITIVE LOGITS

 evenly

0.77

zzle

0.73

 remaining

0.73

lid

0.72

 excess

0.71

 vigorously

0.69

 mixture

0.68

semble

0.68

 gently

0.67

 carefully

0.66

Activations Density 0.099%

instructions or steps in a recipe

cooking actions and instructions

No Comments

No Known Activations