INDEX

Explanations

adjectives relating to unexpected or surprising situations

oai_token-act-pair · gpt-3.5-turbo

terms expressing unexpectedness or peculiarity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ngth

-0.83

rive

-0.69

cipl

-0.69

 matured

-0.63

llular

-0.63

iver

-0.62

ranean

-0.61

 polymorph

-0.61

 governs

-0.61

eely

-0.61

POSITIVE LOGITS

 considering

1.01

 because

0.83

soType

0.75

 Exodus

0.73

LY

0.71

XD

0.70

 given

0.69

JP

0.67

 Especially

0.67

é¾į

0.66

Activations Density 0.203%

adjectives relating to unexpected or surprising situations

terms expressing unexpectedness or peculiarity

No Comments

No Known Activations