INDEX

Explanations

phrases indicating completion or accomplishment

oai_token-act-pair · gpt-3.5-turbo

instances of the word "done" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

lights

-0.73

ulates

-0.73

olson

-0.71

azines

-0.68

ÏĦ

-0.68

anmar

-0.65

ulative

-0.65

eros

-0.63

 Belief

-0.63

mare

-0.63

POSITIVE LOGITS

pez

1.27

ggie

0.78

chy

0.77

 differently

0.75

 administr

0.71

omething

0.70

 brisk

0.70

 away

0.69

ername

0.68

 poorly

0.66

Activations Density 0.038%

phrases indicating completion or accomplishment

instances of the word "done" in various contexts

No Comments

No Known Activations