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

ADRA

-0.88

cyclop

-0.77

icer

-0.76

agara

-0.74

 submar

-0.72

 blat

-0.72

 pecul

-0.72

igree

-0.71

 brill

-0.70

 todd

-0.70

POSITIVE LOGITS

)]

0.82

 JUSTICE

0.73

wolves

0.73

anooga

0.70

ãĥĻ

0.65

ect

0.63

NV

0.63

 unfocusedRange

0.63

Runtime

0.62

ãĥĩãĤ£

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.