INDEX

Explanations

time-related phrases

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate future actions or plans

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

Definition

-0.75

Liter

-0.74

ritical

-0.73

Roman

-0.73

bian

-0.71

usage

-0.68

Vers

-0.67

thodox

-0.67

catentry

-0.66

ullah

-0.65

POSITIVE LOGITS

 plenty

0.70

anmar

0.68

ï¸ı

0.66

]}

0.65

ATURE

0.63

 Delete

0.62

 tune

0.61

 Pablo

0.60

 deleted

0.60

 Adventures

0.59

Activations Density 0.090%

time-related phrases

phrases that indicate future actions or plans

No Comments

No Known Activations