INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

zza

-0.82

odox

-0.74

nea

-0.74

nai

-0.71

aldi

-0.70

kay

-0.69

 Alto

-0.68

stra

-0.68

amina

-0.67

ati

-0.67

POSITIVE LOGITS

 Trace

0.69

=\"

0.65

 Wick

0.62

 toget

0.62

Â·Â·

0.60

 endif

0.60

 Ended

0.60

 abund

0.58

pload

0.58

 Âł Âł Âł Âł

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.