INDEX

Explanations

alphanumeric symbols occurring at the end of a line

oai_token-act-pair · gpt-3.5-turbo

symbols or markers typically used in lists or annotations

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

inator

-0.77

unia

-0.77

ivating

-0.70

ivated

-0.70

seless

-0.69

 force

-0.67

selage

-0.66

 fireplace

-0.65

 cultivated

-0.63

 wisdom

-0.63

POSITIVE LOGITS

 Shards

0.85

(*

0.79

TON

0.78

âĢł

0.76

bah

0.76

AUT

0.76

FREE

0.75

(*

0.75

Thompson

0.74

/-

0.73

Activations Density 0.016%

alphanumeric symbols occurring at the end of a line

symbols or markers typically used in lists or annotations

No Comments

No Known Activations