INDEX

Explanations

words related to actions or processes with a significant impact or consequence

oai_token-act-pair · gpt-3.5-turbo

gerunds or participles that indicate ongoing actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

behind

-0.74

si

-0.68

spot

-0.67

omet

-0.66

adem

-0.66

ande

-0.66

saf

-0.66

nin

-0.63

tw

-0.63

anya

-0.62

POSITIVE LOGITS

redients

0.90

LY

0.86

ulate

0.77

 oneself

0.70

OPLE

0.67

 schedules

0.66

ATIONS

0.66

ly

0.65

enance

0.65

utical

0.65

Activations Density 0.109%

words related to actions or processes with a significant impact or consequence

gerunds or participles that indicate ongoing actions or processes

No Comments

No Known Activations