INDEX

Explanations

phrases related to the name "Crockett."

oai_token-act-pair · gpt-3.5-turbo

words and names related to specific locations or places

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

 contrace

-0.83

ADRA

-0.77

 Shutterstock

-0.72

BIL

-0.69

ku

-0.65

 brightest

-0.64

 satell

-0.64

 seiz

-0.63

fw

-0.61

iage

-0.60

POSITIVE LOGITS

neck

0.89

bart

0.76

esar

0.76

pton

0.76

berry

0.75

ombat

0.73

Berry

0.72

batch

0.71

houn

0.67

urrencies

0.67

Activations Density 0.145%

phrases related to the name "Crockett."

words and names related to specific locations or places

No Comments

No Known Activations