INDEX

Explanations

occurrences of the word "Shorten" in the text

oai_token-act-pair · gpt-3.5-turbo

references to the name "Short" 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 @ 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

ILLE

-0.89

risome

-0.85

itational

-0.85

igators

-0.74

antine

-0.69

Magikarp

-0.67

ILA

-0.67

berus

-0.66

VERTIS

-0.65

ICAN

-0.65

POSITIVE LOGITS

cuts

1.05

comings

1.00

bread

0.97

ened

0.95

cut

0.95

short

0.88

ener

0.86

ening

0.84

eners

0.84

ness

0.81

Activations Density 0.018%

occurrences of the word "Shorten" in the text

references to the name "Short" in various contexts

No Comments

No Known Activations