INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ickr

-0.70

ppe

-0.68

sein

-0.68

eni

-0.66

hol

-0.66

Rating

-0.65

soType

-0.65

facebook

-0.65

yi

-0.64

ocity

-0.64

POSITIVE LOGITS

 candidates

1.13

 candidate

1.08

 Candidate

0.88

 candidacy

0.75

 specialization

0.68

 millenn

0.66

igham

0.65

 Trainer

0.65

 vying

0.65

 nomination

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.