INDEX

Explanations

emblems or symbols within the text

oai_token-act-pair · gpt-3.5-turbo

terms related to diplomatic contexts and symbols

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

 Hath

-0.70

rooms

-0.67

Eck

-0.65

 bras

-0.65

ruff

-0.63

 Berry

-0.63

 Tech

-0.63

room

-0.62

ibr

-0.62

 Palin

-0.61

POSITIVE LOGITS

atic

1.38

atically

1.17

ģĸ

1.01

ĪĴ

0.97

unal

0.96

atis

0.96

alties

0.94

Ħ¢

0.87

Ĥİ

0.87

Į

0.87

Activations Density 0.040%

emblems or symbols within the text

terms related to diplomatic contexts and symbols

No Comments

No Known Activations