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

agine

-0.82

otti

-0.77

 Leone

-0.77

 Aires

-0.72

 interpretations

-0.67

inois

-0.65

Forge

-0.64

 cafes

-0.64

irms

-0.64

Lau

-0.63

POSITIVE LOGITS

gered

0.71

imer

0.68

rote

0.66

ror

0.66

Hug

0.65

 Vald

0.63

phabet

0.62

uliffe

0.62

 poisoned

0.61

ļéĨĴ

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.