INDEX

Explanations

instances of the word "practice"

oai_token-act-pair · gpt-3.5-turbo

references to specific practices or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ilver

-0.68

 Flavoring

-0.67

 Frag

-0.66

Frag

-0.66

semble

-0.64

adow

-0.63

inventoryQuantity

-0.62

incinn

-0.61

orters

-0.61

soon

-0.61

POSITIVE LOGITS

ually

0.90

 whereby

0.88

pmwiki

0.77

 reversal

0.73

Ĥİ

0.70

 gamble

0.69

nels

0.67

ery

0.67

horse

0.67

jam

0.64

Activations Density 0.204%

instances of the word "practice"

references to specific practices or actions

No Comments

No Known Activations