INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Fidel

-0.82

verting

-0.76

stocks

-0.76

verted

-0.74

uala

-0.71

stained

-0.66

ctory

-0.66

audi

-0.64

jer

-0.64

stan

-0.64

POSITIVE LOGITS

\/

0.69

 srfAttach

0.68

KR

0.66

 slate

0.64

 Seah

0.64

eus

0.63

 paws

0.62

 Davidson

0.62

ko

0.61

 Tenn

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.