INDEX

Explanations

negations or exceptions in sentences

oai_token-act-pair · gpt-3.5-turbo

negative constructions and expressions of non-existence or unreality

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

issance

-0.70

ESE

-0.68

ONS

-0.67

 Offline

-0.65

CAN

-0.64

Reviewer

-0.64

soType

-0.64

ON

-0.64

ocard

-0.63

OST

-0.62

POSITIVE LOGITS

 necessarily

1.05

 otherwise

1.04

 ordinarily

0.95

necess

0.90

 previously

0.88

 normally

0.88

fit

0.85

 conform

0.84

 traditionally

0.81

icable

0.77

Activations Density 0.154%

negations or exceptions in sentences

negative constructions and expressions of non-existence or unreality

No Comments

No Known Activations