INDEX

Explanations

It seems that Neuron 4 does not activate for any of the tokens provided, indicating it may be looking for something not present in the given excerpts, or it is not functioning correctly

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

=-=-=-=-=-=-=-=-

-0.75

 classmate

-0.70

 Ping

-0.66

 Patron

-0.65

zza

-0.65

 fres

-0.64

osit

-0.63

 pasta

-0.63

 McCann

-0.63

Pas

-0.61

POSITIVE LOGITS

mson

0.82

ModLoader

0.78

éĹĺ

0.78

iaries

0.74

KN

0.68

)=(

0.68

Kings

0.67

Southern

0.67

ividual

0.66

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.