INDEX

Explanations

words related to names and surnames

oai_token-act-pair · gpt-3.5-turbo

words containing the letter combination "en"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

WTC

-0.67

 benign

-0.66

 hostile

-0.66

EStream

-0.65

 boycot

-0.65

 Hurricanes

-0.64

 polarization

-0.63

TAM

-0.62

DAQ

-0.62

 kindred

-0.61

POSITIVE LOGITS

ake

0.88

inite

0.80

ridge

0.79

hoff

0.77

arten

0.76

ivari

0.75

otide

0.75

Pic

0.75

ash

0.75

aru

0.74

Activations Density 0.214%

words related to names and surnames

words containing the letter combination "en"

No Comments

No Known Activations