INDEX

Explanations

It appears that Neuron 4 did not activate for any of the words provided in the document segments; therefore, no pattern of interest can be identified based on the given information

oai_token-act-pair · gpt-4-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Pitt

-0.71

————————————————

-0.65

 Worcester

-0.65

fall

-0.64

 Hallow

-0.62

Gaza

-0.60

violent

-0.60

00000000

-0.59

Buff

-0.59

 MÃ¼

-0.58

POSITIVE LOGITS

ibaba

0.96

llular

0.78

llah

0.75

orney

0.74

kinson

0.73

ynthesis

0.73

inki

0.72

ilton

0.67

velength

0.66

oya

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.